eandb_test.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "eandb_test.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "toc_visible": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "sQY7n0q4xapg",
        "colab_type": "text"
      },
      "source": [
        "# init"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "xYBpZnF0kcTa",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        },
        "outputId": "a5e07d3f-aeb5-45ed-b3db-5f6d00cfb882"
      },
      "source": [
        "from google.colab import drive\n",
        "drive.mount('/gdrive')\n",
        "\n",
        "%cd /gdrive/My Drive/root ap/Epoch and Batches"
      ],
      "execution_count": 1,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Drive already mounted at /gdrive; to attempt to forcibly remount, call drive.mount(\"/gdrive\", force_remount=True).\n",
            "/gdrive/My Drive/root ap/Epoch and Batches\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "AQwbRGKHwjUQ",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import sys\n",
        "import pandas as pd\n",
        "import matplotlib\n",
        "import matplotlib.pyplot as plt\n",
        "import random\n",
        "import numpy as np\n",
        "import torch\n",
        "from pprint import pprint as pprint"
      ],
      "execution_count": 2,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "PaQ1N6tbzU58",
        "colab_type": "text"
      },
      "source": [
        "# EnB"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "0wQIlhtXy6nr",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from EpochAndBatches import EpochAndBatches as EnB "
      ],
      "execution_count": 3,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "1KOFyUN3zkbL",
        "colab_type": "text"
      },
      "source": [
        "# unit test"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "bSEhEzI0xcpS",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "config = {}\n",
        "config['n_epoch'] = 20\n",
        "config['batch_size'] = 4096\n",
        "config['n_batch'] = 30\n",
        "config['device'] = 'cuda:0'\n",
        "d = config['device']"
      ],
      "execution_count": 4,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "xcKcb45t5Jpy",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 391
        },
        "outputId": "8a220a23-097c-464d-e62d-4062b24709e4"
      },
      "source": [
        "enb = EnB(config['n_epoch'],config['n_batch'])\n",
        "enb.config = config\n",
        "# enb.verbose = 9000\n",
        "\n",
        "enb.save()\n",
        "\n",
        "pprint(vars(enb))"
      ],
      "execution_count": 5,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "EnB | msg = EnB: Epoch And Batches START\n",
            "{'config': {'batch_size': 4096,\n",
            "            'device': 'cuda:0',\n",
            "            'n_batch': 30,\n",
            "            'n_epoch': 20},\n",
            " 'cwd': '/gdrive/My Drive/root ap/Epoch and Batches',\n",
            " 'dflog':    i_batch  i_epoch  key label                         value\n",
            "0     -1.0     -1.0  msg   msg  EnB: Epoch And Batches START,\n",
            " 'enb_output_filepath': 'default_run_1600324884_694.enb',\n",
            " 'i_batch': -1,\n",
            " 'i_epoch': -1,\n",
            " 'is_save_per_batch': False,\n",
            " 'is_save_per_epoch': True,\n",
            " 'is_save_per_log': False,\n",
            " 'n_batch': 30,\n",
            " 'n_epoch': 20,\n",
            " 'pid': 694,\n",
            " 'project_name': 'Default Project',\n",
            " 'run_name': 'default_run_1600324884_694',\n",
            " 'script_start_time': 1600324884.2259302,\n",
            " 'sweep_name': 'Default Sweep',\n",
            " 'verbose': 15}\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "eRNmqm1dQcQf",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "net = torch.nn.Sequential(torch.nn.Linear(1, 4096),\n",
        "                          torch.nn.ReLU(),\n",
        "                          torch.nn.Linear(4096, 1)).to(d)\n",
        "\n",
        "o = torch.optim.Adam(net.parameters())\n",
        "\n",
        "loss_mae = torch.nn.L1Loss()\n",
        "loss_mse = torch.nn.MSELoss()"
      ],
      "execution_count": 6,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "DBvKY_xCzzQB",
        "colab_type": "text"
      },
      "source": [
        "## fit"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Qx0CIQLgxWHX",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 680
        },
        "outputId": "b0a6f915-b6a8-4751-f38d-b88d0416734b"
      },
      "source": [
        "enb.best_epoch = -1\n",
        "enb.best_val_loss = sys.float_info.max\n",
        "for i_epoch in range(config['n_epoch']):\n",
        "    enb.next_epoch()\n",
        "    z_mae = 0\n",
        "    z_mse = 0\n",
        "    zmetric1 = 0\n",
        "    zmetric2 = 0\n",
        "    # train one epoch\n",
        "    for i_batch in range(config['n_batch']):\n",
        "        enb.next_batch()\n",
        "        # forward\n",
        "        net.train()\n",
        "        o.zero_grad()\n",
        "        ix = torch.rand(config['batch_size'],1).to(d) * 4 - 2\n",
        "        iy = ix**2\n",
        "        ih = net(ix)\n",
        "        mae = loss_mae(ih,iy)\n",
        "        mse = loss_mse(ih,iy)\n",
        "        loss = mae**2 + mse\n",
        "        # backward\n",
        "        loss.backward()\n",
        "        o.step()\n",
        "        # metrics\n",
        "        metric1 = mae.item() + mse.item()\n",
        "        metric2 = mae.item() - mse.item()\n",
        "        z_mae += mae.item()\n",
        "        z_mse += mse.item()\n",
        "        zmetric1 += metric1\n",
        "        zmetric2 += metric2\n",
        "        #enb\n",
        "        enb.log_batch('loss:mae',mae.item(),35)\n",
        "        enb.log_batch('loss:mse',mse.item(),35)\n",
        "        enb.log_batch('metric:1',metric1,35)\n",
        "        enb.log_batch('metric:2',metric2,35)\n",
        "    z_mae /= config['n_batch']\n",
        "    z_mse /= config['n_batch']\n",
        "    zmetric1 /= config['n_batch']\n",
        "    zmetric2 /= config['n_batch']\n",
        "    enb.log_epoch('loss:mae',z_mae,35)\n",
        "    enb.log_epoch('loss:mse',z_mse,35)\n",
        "    enb.log_epoch('metric:1',zmetric1,35)\n",
        "    enb.log_epoch('metric:2',zmetric2,35)\n",
        "    enb.log_epoch('gpu-mem',str(torch.cuda.memory_reserved(d)),35)\n",
        "    # validate one epoch\n",
        "    net.eval()\n",
        "    ix = torch.rand(config['batch_size']*5,1).to(d) *6 - 3\n",
        "    iy = ix**2\n",
        "    ih = net(ix)\n",
        "    val_mae = loss_mae(ih,iy).item()\n",
        "    val_mse = loss_mse(ih,iy).item()\n",
        "    loss = val_mae**2 + val_mse\n",
        "    val_metric1 = val_mae + val_mse\n",
        "    val_metric2 = val_mae - val_mse\n",
        "    enb.log_epoch('loss:val_mae',val_mae,35)\n",
        "    enb.log_epoch('loss:val_mse',val_mse,35)\n",
        "    enb.log_epoch('metric:1',val_metric1,35)\n",
        "    enb.log_epoch('metric:2',val_metric2,35)\n",
        "    enb.log_epoch('model:net_state_dict',net.state_dict(),35)\n",
        "    enb.log_epoch('optim:adam_state_dict',o.state_dict(),35)\n",
        "    if loss < enb.best_val_loss:\n",
        "        enb.best_val_loss = loss\n",
        "        enb.best_epoch = enb.i_epoch\n",
        "        enb.best_model_state_dict = net.state_dict()\n",
        "        enb.log_msg('New best_val_loss='+str(enb.best_val_loss))\n",
        "        enb.best_val_mae = val_mae\n",
        "        enb.best_val_mse = val_mse\n",
        "        enb.best_val_metric1 = val_metric1\n",
        "        enb.best_val_metric2 = val_metric2\n",
        "#test\n",
        "net.load_state_dict(enb.best_model_state_dict)\n",
        "ix = torch.rand(config['batch_size']*5,1).to(d) * 8 - 4\n",
        "iy = ix**2\n",
        "ih = net(ix)\n",
        "enb.tst_mae = loss_mae(ih,iy).item()\n",
        "enb.tst_mse = loss_mse(ih,iy).item()\n",
        "enb.tst_metric1 = enb.tst_mae + enb.tst_mse\n",
        "enb.tst_metric2 = enb.tst_mae - enb.tst_mse\n",
        "enb.save()"
      ],
      "execution_count": 7,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "EnB | msg = next_epoch: 0/20  |  00:00:00 + 00:00:00 = 00:00:00 / 00:00:00\n",
            "EnB | msg = New best_val_loss=0.6156449827258257\n",
            "EnB | msg = next_epoch: 1/20  |  00:00:01 + 00:00:20 = 00:00:01 / 00:00:21\n",
            "EnB | msg = New best_val_loss=0.4570068904295903\n",
            "EnB | msg = next_epoch: 2/20  |  00:00:02 + 00:00:19 = 00:00:01 / 00:00:21\n",
            "EnB | msg = New best_val_loss=0.3574326042167053\n",
            "EnB | msg = next_epoch: 3/20  |  00:00:03 + 00:00:18 = 00:00:01 / 00:00:21\n",
            "EnB | msg = New best_val_loss=0.2825830862922827\n",
            "EnB | msg = next_epoch: 4/20  |  00:00:04 + 00:00:17 = 00:00:01 / 00:00:21\n",
            "EnB | msg = New best_val_loss=0.24570834631221516\n",
            "EnB | msg = next_epoch: 5/20  |  00:00:05 + 00:00:16 = 00:00:01 / 00:00:21\n",
            "EnB | msg = New best_val_loss=0.23792475520781609\n",
            "EnB | msg = next_epoch: 6/20  |  00:00:06 + 00:00:15 = 00:00:01 / 00:00:21\n",
            "EnB | msg = New best_val_loss=0.2184428196013235\n",
            "EnB | msg = next_epoch: 7/20  |  00:00:07 + 00:00:14 = 00:00:01 / 00:00:21\n",
            "EnB | msg = New best_val_loss=0.20539241449972523\n",
            "EnB | msg = next_epoch: 8/20  |  00:00:08 + 00:00:12 = 00:00:01 / 00:00:21\n",
            "EnB | msg = next_epoch: 9/20  |  00:00:09 + 00:00:11 = 00:00:01 / 00:00:21\n",
            "EnB | msg = New best_val_loss=0.20438469368557421\n",
            "EnB | msg = next_epoch: 10/20  |  00:00:10 + 00:00:10 = 00:00:01 / 00:00:21\n",
            "EnB | msg = New best_val_loss=0.19485849443368863\n",
            "EnB | msg = next_epoch: 11/20  |  00:00:11 + 00:00:09 = 00:00:01 / 00:00:21\n",
            "EnB | msg = New best_val_loss=0.19069741274290752\n",
            "EnB | msg = next_epoch: 12/20  |  00:00:13 + 00:00:08 = 00:00:01 / 00:00:21\n",
            "EnB | msg = New best_val_loss=0.18494376061742757\n",
            "EnB | msg = next_epoch: 13/20  |  00:00:14 + 00:00:07 = 00:00:01 / 00:00:21\n",
            "EnB | msg = New best_val_loss=0.18277202030900797\n",
            "EnB | msg = next_epoch: 14/20  |  00:00:15 + 00:00:06 = 00:00:01 / 00:00:21\n",
            "EnB | msg = New best_val_loss=0.18129668916691188\n",
            "EnB | msg = next_epoch: 15/20  |  00:00:16 + 00:00:05 = 00:00:01 / 00:00:22\n",
            "EnB | msg = New best_val_loss=0.1797702233965064\n",
            "EnB | msg = next_epoch: 16/20  |  00:00:17 + 00:00:04 = 00:00:01 / 00:00:22\n",
            "EnB | msg = New best_val_loss=0.1784326787639705\n",
            "EnB | msg = next_epoch: 17/20  |  00:00:18 + 00:00:03 = 00:00:01 / 00:00:22\n",
            "EnB | msg = New best_val_loss=0.17438779729287668\n",
            "EnB | msg = next_epoch: 18/20  |  00:00:19 + 00:00:02 = 00:00:01 / 00:00:22\n",
            "EnB | msg = New best_val_loss=0.1713978835977663\n",
            "EnB | msg = next_epoch: 19/20  |  00:00:21 + 00:00:01 = 00:00:01 / 00:00:22\n",
            "EnB | msg = New best_val_loss=0.16876476377409433\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "b1B_Un53qHlz",
        "colab_type": "text"
      },
      "source": [
        "# load"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ms0usMNaqJ5a",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        },
        "outputId": "faa6f438-04ef-461d-ee79-f3709a2b0c2e"
      },
      "source": [
        "enb2 = EnB.load(enb.enb_output_filepath)"
      ],
      "execution_count": 8,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "EnB loading  default_run_1600324884_694.enb  ...\n",
            "EnB  default_run_1600324884_694.enb  loaded.\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ivStuDh7ulJ-",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 527
        },
        "outputId": "a64ba24f-db63-46e4-eed4-d58aa449a976"
      },
      "source": [
        "d_enb = vars(enb2)\n",
        "for k,v in d_enb.items():\n",
        "    print(k,type(v))"
      ],
      "execution_count": 16,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "project_name <class 'str'>\n",
            "sweep_name <class 'str'>\n",
            "dflog <class 'pandas.core.frame.DataFrame'>\n",
            "script_start_time <class 'float'>\n",
            "pid <class 'int'>\n",
            "cwd <class 'str'>\n",
            "verbose <class 'int'>\n",
            "n_epoch <class 'int'>\n",
            "n_batch <class 'int'>\n",
            "i_epoch <class 'int'>\n",
            "i_batch <class 'int'>\n",
            "is_save_per_log <class 'bool'>\n",
            "is_save_per_batch <class 'bool'>\n",
            "is_save_per_epoch <class 'bool'>\n",
            "run_name <class 'str'>\n",
            "enb_output_filepath <class 'str'>\n",
            "config <class 'dict'>\n",
            "best_epoch <class 'int'>\n",
            "best_val_loss <class 'float'>\n",
            "fit_start_time <class 'float'>\n",
            "i_batch_start_time <class 'float'>\n",
            "best_model_state_dict <class 'collections.OrderedDict'>\n",
            "best_val_mae <class 'float'>\n",
            "best_val_mse <class 'float'>\n",
            "best_val_metric1 <class 'float'>\n",
            "best_val_metric2 <class 'float'>\n",
            "tst_mae <class 'float'>\n",
            "tst_mse <class 'float'>\n",
            "tst_metric1 <class 'float'>\n",
            "tst_metric2 <class 'float'>\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "pcZwXf_9xWFE",
        "colab_type": "text"
      },
      "source": [
        "# display"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "1dRgwSsQxRXN",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 391
        },
        "outputId": "887d482d-a57d-4064-a773-1a03466f488d"
      },
      "source": [
        "for il in enb2.dflog['label'].unique():\n",
        "    for ik in enb2.dflog['key'].loc[enb2.dflog['label']==il].unique():\n",
        "        print(il,ik)"
      ],
      "execution_count": 10,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "msg msg\n",
            "per_epoch time:elapsed\n",
            "per_epoch time:total\n",
            "per_epoch time:left\n",
            "per_epoch time:each\n",
            "per_epoch loss:mae\n",
            "per_epoch loss:mse\n",
            "per_epoch metric:1\n",
            "per_epoch metric:2\n",
            "per_epoch gpu-mem\n",
            "per_epoch loss:val_mae\n",
            "per_epoch loss:val_mse\n",
            "per_epoch model:net_state_dict\n",
            "per_epoch optim:adam_state_dict\n",
            "per_batch time:elapsed\n",
            "per_batch time:total\n",
            "per_batch time:left\n",
            "per_batch time:each\n",
            "per_batch loss:mae\n",
            "per_batch loss:mse\n",
            "per_batch metric:1\n",
            "per_batch metric:2\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "3CWjnwc0rmHo",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "d6809abe-d83c-4320-fd27-62d9c820a7c1"
      },
      "source": [
        "for il in enb2.dflog['label'].unique():\n",
        "    for ik in enb2.dflog['key'].loc[enb2.dflog['label']==il].unique():\n",
        "        print(il,ik)\n",
        "        display(enb2.dflog.loc[(enb2.dflog['key']==ik) & (enb2.dflog['label']==il)])"
      ],
      "execution_count": 11,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "msg msg\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>i_batch</th>\n",
              "      <th>i_epoch</th>\n",
              "      <th>key</th>\n",
              "      <th>label</th>\n",
              "      <th>value</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>-1.0</td>\n",
              "      <td>-1.0</td>\n",
              "      <td>msg</td>\n",
              "      <td>msg</td>\n",
              "      <td>EnB: Epoch And Batches START</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>-1.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>msg</td>\n",
              "      <td>msg</td>\n",
              "      <td>next_epoch: 0/20  |  00:00:00 + 00:00:00 = 00:...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>257</th>\n",
              "      <td>29.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>msg</td>\n",
              "      <td>msg</td>\n",
              "      <td>New best_val_loss=0.6156449827258257</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>258</th>\n",
              "      <td>-1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>msg</td>\n",
              "      <td>msg</td>\n",
              "      <td>next_epoch: 1/20  |  00:00:01 + 00:00:20 = 00:...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>514</th>\n",
              "      <td>29.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>msg</td>\n",
              "      <td>msg</td>\n",
              "      <td>New best_val_loss=0.4570068904295903</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>515</th>\n",
              "      <td>-1.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>msg</td>\n",
              "      <td>msg</td>\n",
              "      <td>next_epoch: 2/20  |  00:00:02 + 00:00:19 = 00:...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>771</th>\n",
              "      <td>29.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>msg</td>\n",
              "      <td>msg</td>\n",
              "      <td>New best_val_loss=0.3574326042167053</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>772</th>\n",
              "      <td>-1.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>msg</td>\n",
              "      <td>msg</td>\n",
              "      <td>next_epoch: 3/20  |  00:00:03 + 00:00:18 = 00:...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1028</th>\n",
              "      <td>29.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>msg</td>\n",
              "      <td>msg</td>\n",
              "      <td>New best_val_loss=0.2825830862922827</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1029</th>\n",
              "      <td>-1.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>msg</td>\n",
              "      <td>msg</td>\n",
              "      <td>next_epoch: 4/20  |  00:00:04 + 00:00:17 = 00:...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1285</th>\n",
              "      <td>29.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>msg</td>\n",
              "      <td>msg</td>\n",
              "      <td>New best_val_loss=0.24570834631221516</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1286</th>\n",
              "      <td>-1.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>msg</td>\n",
              "      <td>msg</td>\n",
              "      <td>next_epoch: 5/20  |  00:00:05 + 00:00:16 = 00:...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1542</th>\n",
              "      <td>29.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>msg</td>\n",
              "      <td>msg</td>\n",
              "      <td>New best_val_loss=0.23792475520781609</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1543</th>\n",
              "      <td>-1.0</td>\n",
              "      <td>6.0</td>\n",
              "      <td>msg</td>\n",
              "      <td>msg</td>\n",
              "      <td>next_epoch: 6/20  |  00:00:06 + 00:00:15 = 00:...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1799</th>\n",
              "      <td>29.0</td>\n",
              "      <td>6.0</td>\n",
              "      <td>msg</td>\n",
              "      <td>msg</td>\n",
              "      <td>New best_val_loss=0.2184428196013235</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1800</th>\n",
              "      <td>-1.0</td>\n",
              "      <td>7.0</td>\n",
              "      <td>msg</td>\n",
              "      <td>msg</td>\n",
              "      <td>next_epoch: 7/20  |  00:00:07 + 00:00:14 = 00:...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2056</th>\n",
              "      <td>29.0</td>\n",
              "      <td>7.0</td>\n",
              "      <td>msg</td>\n",
              "      <td>msg</td>\n",
              "      <td>New best_val_loss=0.20539241449972523</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2057</th>\n",
              "      <td>-1.0</td>\n",
              "      <td>8.0</td>\n",
              "      <td>msg</td>\n",
              "      <td>msg</td>\n",
              "      <td>next_epoch: 8/20  |  00:00:08 + 00:00:12 = 00:...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2313</th>\n",
              "      <td>-1.0</td>\n",
              "      <td>9.0</td>\n",
              "      <td>msg</td>\n",
              "      <td>msg</td>\n",
              "      <td>next_epoch: 9/20  |  00:00:09 + 00:00:11 = 00:...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2569</th>\n",
              "      <td>29.0</td>\n",
              "      <td>9.0</td>\n",
              "      <td>msg</td>\n",
              "      <td>msg</td>\n",
              "      <td>New best_val_loss=0.20438469368557421</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2570</th>\n",
              "      <td>-1.0</td>\n",
              "      <td>10.0</td>\n",
              "      <td>msg</td>\n",
              "      <td>msg</td>\n",
              "      <td>next_epoch: 10/20  |  00:00:10 + 00:00:10 = 00...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2826</th>\n",
              "      <td>29.0</td>\n",
              "      <td>10.0</td>\n",
              "      <td>msg</td>\n",
              "      <td>msg</td>\n",
              "      <td>New best_val_loss=0.19485849443368863</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2827</th>\n",
              "      <td>-1.0</td>\n",
              "      <td>11.0</td>\n",
              "      <td>msg</td>\n",
              "      <td>msg</td>\n",
              "      <td>next_epoch: 11/20  |  00:00:11 + 00:00:09 = 00...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3083</th>\n",
              "      <td>29.0</td>\n",
              "      <td>11.0</td>\n",
              "      <td>msg</td>\n",
              "      <td>msg</td>\n",
              "      <td>New best_val_loss=0.19069741274290752</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3084</th>\n",
              "      <td>-1.0</td>\n",
              "      <td>12.0</td>\n",
              "      <td>msg</td>\n",
              "      <td>msg</td>\n",
              "      <td>next_epoch: 12/20  |  00:00:13 + 00:00:08 = 00...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3340</th>\n",
              "      <td>29.0</td>\n",
              "      <td>12.0</td>\n",
              "      <td>msg</td>\n",
              "      <td>msg</td>\n",
              "      <td>New best_val_loss=0.18494376061742757</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3341</th>\n",
              "      <td>-1.0</td>\n",
              "      <td>13.0</td>\n",
              "      <td>msg</td>\n",
              "      <td>msg</td>\n",
              "      <td>next_epoch: 13/20  |  00:00:14 + 00:00:07 = 00...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3597</th>\n",
              "      <td>29.0</td>\n",
              "      <td>13.0</td>\n",
              "      <td>msg</td>\n",
              "      <td>msg</td>\n",
              "      <td>New best_val_loss=0.18277202030900797</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3598</th>\n",
              "      <td>-1.0</td>\n",
              "      <td>14.0</td>\n",
              "      <td>msg</td>\n",
              "      <td>msg</td>\n",
              "      <td>next_epoch: 14/20  |  00:00:15 + 00:00:06 = 00...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3854</th>\n",
              "      <td>29.0</td>\n",
              "      <td>14.0</td>\n",
              "      <td>msg</td>\n",
              "      <td>msg</td>\n",
              "      <td>New best_val_loss=0.18129668916691188</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3855</th>\n",
              "      <td>-1.0</td>\n",
              "      <td>15.0</td>\n",
              "      <td>msg</td>\n",
              "      <td>msg</td>\n",
              "      <td>next_epoch: 15/20  |  00:00:16 + 00:00:05 = 00...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4111</th>\n",
              "      <td>29.0</td>\n",
              "      <td>15.0</td>\n",
              "      <td>msg</td>\n",
              "      <td>msg</td>\n",
              "      <td>New best_val_loss=0.1797702233965064</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4112</th>\n",
              "      <td>-1.0</td>\n",
              "      <td>16.0</td>\n",
              "      <td>msg</td>\n",
              "      <td>msg</td>\n",
              "      <td>next_epoch: 16/20  |  00:00:17 + 00:00:04 = 00...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4368</th>\n",
              "      <td>29.0</td>\n",
              "      <td>16.0</td>\n",
              "      <td>msg</td>\n",
              "      <td>msg</td>\n",
              "      <td>New best_val_loss=0.1784326787639705</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4369</th>\n",
              "      <td>-1.0</td>\n",
              "      <td>17.0</td>\n",
              "      <td>msg</td>\n",
              "      <td>msg</td>\n",
              "      <td>next_epoch: 17/20  |  00:00:18 + 00:00:03 = 00...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4625</th>\n",
              "      <td>29.0</td>\n",
              "      <td>17.0</td>\n",
              "      <td>msg</td>\n",
              "      <td>msg</td>\n",
              "      <td>New best_val_loss=0.17438779729287668</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4626</th>\n",
              "      <td>-1.0</td>\n",
              "      <td>18.0</td>\n",
              "      <td>msg</td>\n",
              "      <td>msg</td>\n",
              "      <td>next_epoch: 18/20  |  00:00:19 + 00:00:02 = 00...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4882</th>\n",
              "      <td>29.0</td>\n",
              "      <td>18.0</td>\n",
              "      <td>msg</td>\n",
              "      <td>msg</td>\n",
              "      <td>New best_val_loss=0.1713978835977663</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4883</th>\n",
              "      <td>-1.0</td>\n",
              "      <td>19.0</td>\n",
              "      <td>msg</td>\n",
              "      <td>msg</td>\n",
              "      <td>next_epoch: 19/20  |  00:00:21 + 00:00:01 = 00...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5139</th>\n",
              "      <td>29.0</td>\n",
              "      <td>19.0</td>\n",
              "      <td>msg</td>\n",
              "      <td>msg</td>\n",
              "      <td>New best_val_loss=0.16876476377409433</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "      i_batch  i_epoch  ... label                                              value\n",
              "0        -1.0     -1.0  ...   msg                       EnB: Epoch And Batches START\n",
              "1        -1.0      0.0  ...   msg  next_epoch: 0/20  |  00:00:00 + 00:00:00 = 00:...\n",
              "257      29.0      0.0  ...   msg               New best_val_loss=0.6156449827258257\n",
              "258      -1.0      1.0  ...   msg  next_epoch: 1/20  |  00:00:01 + 00:00:20 = 00:...\n",
              "514      29.0      1.0  ...   msg               New best_val_loss=0.4570068904295903\n",
              "515      -1.0      2.0  ...   msg  next_epoch: 2/20  |  00:00:02 + 00:00:19 = 00:...\n",
              "771      29.0      2.0  ...   msg               New best_val_loss=0.3574326042167053\n",
              "772      -1.0      3.0  ...   msg  next_epoch: 3/20  |  00:00:03 + 00:00:18 = 00:...\n",
              "1028     29.0      3.0  ...   msg               New best_val_loss=0.2825830862922827\n",
              "1029     -1.0      4.0  ...   msg  next_epoch: 4/20  |  00:00:04 + 00:00:17 = 00:...\n",
              "1285     29.0      4.0  ...   msg              New best_val_loss=0.24570834631221516\n",
              "1286     -1.0      5.0  ...   msg  next_epoch: 5/20  |  00:00:05 + 00:00:16 = 00:...\n",
              "1542     29.0      5.0  ...   msg              New best_val_loss=0.23792475520781609\n",
              "1543     -1.0      6.0  ...   msg  next_epoch: 6/20  |  00:00:06 + 00:00:15 = 00:...\n",
              "1799     29.0      6.0  ...   msg               New best_val_loss=0.2184428196013235\n",
              "1800     -1.0      7.0  ...   msg  next_epoch: 7/20  |  00:00:07 + 00:00:14 = 00:...\n",
              "2056     29.0      7.0  ...   msg              New best_val_loss=0.20539241449972523\n",
              "2057     -1.0      8.0  ...   msg  next_epoch: 8/20  |  00:00:08 + 00:00:12 = 00:...\n",
              "2313     -1.0      9.0  ...   msg  next_epoch: 9/20  |  00:00:09 + 00:00:11 = 00:...\n",
              "2569     29.0      9.0  ...   msg              New best_val_loss=0.20438469368557421\n",
              "2570     -1.0     10.0  ...   msg  next_epoch: 10/20  |  00:00:10 + 00:00:10 = 00...\n",
              "2826     29.0     10.0  ...   msg              New best_val_loss=0.19485849443368863\n",
              "2827     -1.0     11.0  ...   msg  next_epoch: 11/20  |  00:00:11 + 00:00:09 = 00...\n",
              "3083     29.0     11.0  ...   msg              New best_val_loss=0.19069741274290752\n",
              "3084     -1.0     12.0  ...   msg  next_epoch: 12/20  |  00:00:13 + 00:00:08 = 00...\n",
              "3340     29.0     12.0  ...   msg              New best_val_loss=0.18494376061742757\n",
              "3341     -1.0     13.0  ...   msg  next_epoch: 13/20  |  00:00:14 + 00:00:07 = 00...\n",
              "3597     29.0     13.0  ...   msg              New best_val_loss=0.18277202030900797\n",
              "3598     -1.0     14.0  ...   msg  next_epoch: 14/20  |  00:00:15 + 00:00:06 = 00...\n",
              "3854     29.0     14.0  ...   msg              New best_val_loss=0.18129668916691188\n",
              "3855     -1.0     15.0  ...   msg  next_epoch: 15/20  |  00:00:16 + 00:00:05 = 00...\n",
              "4111     29.0     15.0  ...   msg               New best_val_loss=0.1797702233965064\n",
              "4112     -1.0     16.0  ...   msg  next_epoch: 16/20  |  00:00:17 + 00:00:04 = 00...\n",
              "4368     29.0     16.0  ...   msg               New best_val_loss=0.1784326787639705\n",
              "4369     -1.0     17.0  ...   msg  next_epoch: 17/20  |  00:00:18 + 00:00:03 = 00...\n",
              "4625     29.0     17.0  ...   msg              New best_val_loss=0.17438779729287668\n",
              "4626     -1.0     18.0  ...   msg  next_epoch: 18/20  |  00:00:19 + 00:00:02 = 00...\n",
              "4882     29.0     18.0  ...   msg               New best_val_loss=0.1713978835977663\n",
              "4883     -1.0     19.0  ...   msg  next_epoch: 19/20  |  00:00:21 + 00:00:01 = 00...\n",
              "5139     29.0     19.0  ...   msg              New best_val_loss=0.16876476377409433\n",
              "\n",
              "[40 rows x 5 columns]"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "per_epoch time:elapsed\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>i_batch</th>\n",
              "      <th>i_epoch</th>\n",
              "      <th>key</th>\n",
              "      <th>label</th>\n",
              "      <th>value</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>NaN</td>\n",
              "      <td>0.0</td>\n",
              "      <td>time:elapsed</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>259</th>\n",
              "      <td>NaN</td>\n",
              "      <td>1.0</td>\n",
              "      <td>time:elapsed</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>1.0845</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>516</th>\n",
              "      <td>NaN</td>\n",
              "      <td>2.0</td>\n",
              "      <td>time:elapsed</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>2.15855</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>773</th>\n",
              "      <td>NaN</td>\n",
              "      <td>3.0</td>\n",
              "      <td>time:elapsed</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>3.22961</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1030</th>\n",
              "      <td>NaN</td>\n",
              "      <td>4.0</td>\n",
              "      <td>time:elapsed</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>4.30641</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1287</th>\n",
              "      <td>NaN</td>\n",
              "      <td>5.0</td>\n",
              "      <td>time:elapsed</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>5.38274</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1544</th>\n",
              "      <td>NaN</td>\n",
              "      <td>6.0</td>\n",
              "      <td>time:elapsed</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>6.46934</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1801</th>\n",
              "      <td>NaN</td>\n",
              "      <td>7.0</td>\n",
              "      <td>time:elapsed</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>7.54712</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2058</th>\n",
              "      <td>NaN</td>\n",
              "      <td>8.0</td>\n",
              "      <td>time:elapsed</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>8.62646</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2314</th>\n",
              "      <td>NaN</td>\n",
              "      <td>9.0</td>\n",
              "      <td>time:elapsed</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>9.70078</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2571</th>\n",
              "      <td>NaN</td>\n",
              "      <td>10.0</td>\n",
              "      <td>time:elapsed</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>10.8191</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2828</th>\n",
              "      <td>NaN</td>\n",
              "      <td>11.0</td>\n",
              "      <td>time:elapsed</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>11.9238</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3085</th>\n",
              "      <td>NaN</td>\n",
              "      <td>12.0</td>\n",
              "      <td>time:elapsed</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>13.0351</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3342</th>\n",
              "      <td>NaN</td>\n",
              "      <td>13.0</td>\n",
              "      <td>time:elapsed</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>14.16</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3599</th>\n",
              "      <td>NaN</td>\n",
              "      <td>14.0</td>\n",
              "      <td>time:elapsed</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>15.2995</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3856</th>\n",
              "      <td>NaN</td>\n",
              "      <td>15.0</td>\n",
              "      <td>time:elapsed</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>16.5163</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4113</th>\n",
              "      <td>NaN</td>\n",
              "      <td>16.0</td>\n",
              "      <td>time:elapsed</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>17.6824</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4370</th>\n",
              "      <td>NaN</td>\n",
              "      <td>17.0</td>\n",
              "      <td>time:elapsed</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>18.829</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4627</th>\n",
              "      <td>NaN</td>\n",
              "      <td>18.0</td>\n",
              "      <td>time:elapsed</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>19.9397</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4884</th>\n",
              "      <td>NaN</td>\n",
              "      <td>19.0</td>\n",
              "      <td>time:elapsed</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>21.0808</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "      i_batch  i_epoch           key      label    value\n",
              "2         NaN      0.0  time:elapsed  per_epoch        0\n",
              "259       NaN      1.0  time:elapsed  per_epoch   1.0845\n",
              "516       NaN      2.0  time:elapsed  per_epoch  2.15855\n",
              "773       NaN      3.0  time:elapsed  per_epoch  3.22961\n",
              "1030      NaN      4.0  time:elapsed  per_epoch  4.30641\n",
              "1287      NaN      5.0  time:elapsed  per_epoch  5.38274\n",
              "1544      NaN      6.0  time:elapsed  per_epoch  6.46934\n",
              "1801      NaN      7.0  time:elapsed  per_epoch  7.54712\n",
              "2058      NaN      8.0  time:elapsed  per_epoch  8.62646\n",
              "2314      NaN      9.0  time:elapsed  per_epoch  9.70078\n",
              "2571      NaN     10.0  time:elapsed  per_epoch  10.8191\n",
              "2828      NaN     11.0  time:elapsed  per_epoch  11.9238\n",
              "3085      NaN     12.0  time:elapsed  per_epoch  13.0351\n",
              "3342      NaN     13.0  time:elapsed  per_epoch    14.16\n",
              "3599      NaN     14.0  time:elapsed  per_epoch  15.2995\n",
              "3856      NaN     15.0  time:elapsed  per_epoch  16.5163\n",
              "4113      NaN     16.0  time:elapsed  per_epoch  17.6824\n",
              "4370      NaN     17.0  time:elapsed  per_epoch   18.829\n",
              "4627      NaN     18.0  time:elapsed  per_epoch  19.9397\n",
              "4884      NaN     19.0  time:elapsed  per_epoch  21.0808"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "per_epoch time:total\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>i_batch</th>\n",
              "      <th>i_epoch</th>\n",
              "      <th>key</th>\n",
              "      <th>label</th>\n",
              "      <th>value</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>NaN</td>\n",
              "      <td>0.0</td>\n",
              "      <td>time:total</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>260</th>\n",
              "      <td>NaN</td>\n",
              "      <td>1.0</td>\n",
              "      <td>time:total</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>21.6899</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>517</th>\n",
              "      <td>NaN</td>\n",
              "      <td>2.0</td>\n",
              "      <td>time:total</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>21.5855</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>774</th>\n",
              "      <td>NaN</td>\n",
              "      <td>3.0</td>\n",
              "      <td>time:total</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>21.5307</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1031</th>\n",
              "      <td>NaN</td>\n",
              "      <td>4.0</td>\n",
              "      <td>time:total</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>21.5321</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1288</th>\n",
              "      <td>NaN</td>\n",
              "      <td>5.0</td>\n",
              "      <td>time:total</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>21.531</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1545</th>\n",
              "      <td>NaN</td>\n",
              "      <td>6.0</td>\n",
              "      <td>time:total</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>21.5645</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1802</th>\n",
              "      <td>NaN</td>\n",
              "      <td>7.0</td>\n",
              "      <td>time:total</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>21.5632</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2059</th>\n",
              "      <td>NaN</td>\n",
              "      <td>8.0</td>\n",
              "      <td>time:total</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>21.5661</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2315</th>\n",
              "      <td>NaN</td>\n",
              "      <td>9.0</td>\n",
              "      <td>time:total</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>21.5573</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2572</th>\n",
              "      <td>NaN</td>\n",
              "      <td>10.0</td>\n",
              "      <td>time:total</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>21.6381</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2829</th>\n",
              "      <td>NaN</td>\n",
              "      <td>11.0</td>\n",
              "      <td>time:total</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>21.6796</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3086</th>\n",
              "      <td>NaN</td>\n",
              "      <td>12.0</td>\n",
              "      <td>time:total</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>21.7252</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3343</th>\n",
              "      <td>NaN</td>\n",
              "      <td>13.0</td>\n",
              "      <td>time:total</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>21.7846</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3600</th>\n",
              "      <td>NaN</td>\n",
              "      <td>14.0</td>\n",
              "      <td>time:total</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>21.8564</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3857</th>\n",
              "      <td>NaN</td>\n",
              "      <td>15.0</td>\n",
              "      <td>time:total</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>22.0218</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4114</th>\n",
              "      <td>NaN</td>\n",
              "      <td>16.0</td>\n",
              "      <td>time:total</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>22.103</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4371</th>\n",
              "      <td>NaN</td>\n",
              "      <td>17.0</td>\n",
              "      <td>time:total</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>22.1518</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4628</th>\n",
              "      <td>NaN</td>\n",
              "      <td>18.0</td>\n",
              "      <td>time:total</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>22.1552</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4885</th>\n",
              "      <td>NaN</td>\n",
              "      <td>19.0</td>\n",
              "      <td>time:total</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>22.1903</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "      i_batch  i_epoch         key      label    value\n",
              "3         NaN      0.0  time:total  per_epoch        0\n",
              "260       NaN      1.0  time:total  per_epoch  21.6899\n",
              "517       NaN      2.0  time:total  per_epoch  21.5855\n",
              "774       NaN      3.0  time:total  per_epoch  21.5307\n",
              "1031      NaN      4.0  time:total  per_epoch  21.5321\n",
              "1288      NaN      5.0  time:total  per_epoch   21.531\n",
              "1545      NaN      6.0  time:total  per_epoch  21.5645\n",
              "1802      NaN      7.0  time:total  per_epoch  21.5632\n",
              "2059      NaN      8.0  time:total  per_epoch  21.5661\n",
              "2315      NaN      9.0  time:total  per_epoch  21.5573\n",
              "2572      NaN     10.0  time:total  per_epoch  21.6381\n",
              "2829      NaN     11.0  time:total  per_epoch  21.6796\n",
              "3086      NaN     12.0  time:total  per_epoch  21.7252\n",
              "3343      NaN     13.0  time:total  per_epoch  21.7846\n",
              "3600      NaN     14.0  time:total  per_epoch  21.8564\n",
              "3857      NaN     15.0  time:total  per_epoch  22.0218\n",
              "4114      NaN     16.0  time:total  per_epoch   22.103\n",
              "4371      NaN     17.0  time:total  per_epoch  22.1518\n",
              "4628      NaN     18.0  time:total  per_epoch  22.1552\n",
              "4885      NaN     19.0  time:total  per_epoch  22.1903"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "per_epoch time:left\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>i_batch</th>\n",
              "      <th>i_epoch</th>\n",
              "      <th>key</th>\n",
              "      <th>label</th>\n",
              "      <th>value</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>NaN</td>\n",
              "      <td>0.0</td>\n",
              "      <td>time:left</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>261</th>\n",
              "      <td>NaN</td>\n",
              "      <td>1.0</td>\n",
              "      <td>time:left</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>20.6054</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>518</th>\n",
              "      <td>NaN</td>\n",
              "      <td>2.0</td>\n",
              "      <td>time:left</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>19.427</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>775</th>\n",
              "      <td>NaN</td>\n",
              "      <td>3.0</td>\n",
              "      <td>time:left</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>18.3011</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1032</th>\n",
              "      <td>NaN</td>\n",
              "      <td>4.0</td>\n",
              "      <td>time:left</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>17.2257</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1289</th>\n",
              "      <td>NaN</td>\n",
              "      <td>5.0</td>\n",
              "      <td>time:left</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>16.1482</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1546</th>\n",
              "      <td>NaN</td>\n",
              "      <td>6.0</td>\n",
              "      <td>time:left</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>15.0951</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1803</th>\n",
              "      <td>NaN</td>\n",
              "      <td>7.0</td>\n",
              "      <td>time:left</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>14.0161</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2060</th>\n",
              "      <td>NaN</td>\n",
              "      <td>8.0</td>\n",
              "      <td>time:left</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>12.9397</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2316</th>\n",
              "      <td>NaN</td>\n",
              "      <td>9.0</td>\n",
              "      <td>time:left</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>11.8565</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2573</th>\n",
              "      <td>NaN</td>\n",
              "      <td>10.0</td>\n",
              "      <td>time:left</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>10.8191</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2830</th>\n",
              "      <td>NaN</td>\n",
              "      <td>11.0</td>\n",
              "      <td>time:left</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>9.7558</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3087</th>\n",
              "      <td>NaN</td>\n",
              "      <td>12.0</td>\n",
              "      <td>time:left</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>8.69009</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3344</th>\n",
              "      <td>NaN</td>\n",
              "      <td>13.0</td>\n",
              "      <td>time:left</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>7.62461</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3601</th>\n",
              "      <td>NaN</td>\n",
              "      <td>14.0</td>\n",
              "      <td>time:left</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>6.55693</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3858</th>\n",
              "      <td>NaN</td>\n",
              "      <td>15.0</td>\n",
              "      <td>time:left</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>5.50544</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4115</th>\n",
              "      <td>NaN</td>\n",
              "      <td>16.0</td>\n",
              "      <td>time:left</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>4.42059</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4372</th>\n",
              "      <td>NaN</td>\n",
              "      <td>17.0</td>\n",
              "      <td>time:left</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>3.32277</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4629</th>\n",
              "      <td>NaN</td>\n",
              "      <td>18.0</td>\n",
              "      <td>time:left</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>2.21552</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4886</th>\n",
              "      <td>NaN</td>\n",
              "      <td>19.0</td>\n",
              "      <td>time:left</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>1.10951</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "      i_batch  i_epoch        key      label    value\n",
              "4         NaN      0.0  time:left  per_epoch        0\n",
              "261       NaN      1.0  time:left  per_epoch  20.6054\n",
              "518       NaN      2.0  time:left  per_epoch   19.427\n",
              "775       NaN      3.0  time:left  per_epoch  18.3011\n",
              "1032      NaN      4.0  time:left  per_epoch  17.2257\n",
              "1289      NaN      5.0  time:left  per_epoch  16.1482\n",
              "1546      NaN      6.0  time:left  per_epoch  15.0951\n",
              "1803      NaN      7.0  time:left  per_epoch  14.0161\n",
              "2060      NaN      8.0  time:left  per_epoch  12.9397\n",
              "2316      NaN      9.0  time:left  per_epoch  11.8565\n",
              "2573      NaN     10.0  time:left  per_epoch  10.8191\n",
              "2830      NaN     11.0  time:left  per_epoch   9.7558\n",
              "3087      NaN     12.0  time:left  per_epoch  8.69009\n",
              "3344      NaN     13.0  time:left  per_epoch  7.62461\n",
              "3601      NaN     14.0  time:left  per_epoch  6.55693\n",
              "3858      NaN     15.0  time:left  per_epoch  5.50544\n",
              "4115      NaN     16.0  time:left  per_epoch  4.42059\n",
              "4372      NaN     17.0  time:left  per_epoch  3.32277\n",
              "4629      NaN     18.0  time:left  per_epoch  2.21552\n",
              "4886      NaN     19.0  time:left  per_epoch  1.10951"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "per_epoch time:each\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>i_batch</th>\n",
              "      <th>i_epoch</th>\n",
              "      <th>key</th>\n",
              "      <th>label</th>\n",
              "      <th>value</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>NaN</td>\n",
              "      <td>0.0</td>\n",
              "      <td>time:each</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>262</th>\n",
              "      <td>NaN</td>\n",
              "      <td>1.0</td>\n",
              "      <td>time:each</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>1.0845</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>519</th>\n",
              "      <td>NaN</td>\n",
              "      <td>2.0</td>\n",
              "      <td>time:each</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>1.07928</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>776</th>\n",
              "      <td>NaN</td>\n",
              "      <td>3.0</td>\n",
              "      <td>time:each</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>1.07654</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1033</th>\n",
              "      <td>NaN</td>\n",
              "      <td>4.0</td>\n",
              "      <td>time:each</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>1.0766</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1290</th>\n",
              "      <td>NaN</td>\n",
              "      <td>5.0</td>\n",
              "      <td>time:each</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>1.07655</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1547</th>\n",
              "      <td>NaN</td>\n",
              "      <td>6.0</td>\n",
              "      <td>time:each</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>1.07822</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1804</th>\n",
              "      <td>NaN</td>\n",
              "      <td>7.0</td>\n",
              "      <td>time:each</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>1.07816</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2061</th>\n",
              "      <td>NaN</td>\n",
              "      <td>8.0</td>\n",
              "      <td>time:each</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>1.07831</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2317</th>\n",
              "      <td>NaN</td>\n",
              "      <td>9.0</td>\n",
              "      <td>time:each</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>1.07786</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2574</th>\n",
              "      <td>NaN</td>\n",
              "      <td>10.0</td>\n",
              "      <td>time:each</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>1.08191</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2831</th>\n",
              "      <td>NaN</td>\n",
              "      <td>11.0</td>\n",
              "      <td>time:each</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>1.08398</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3088</th>\n",
              "      <td>NaN</td>\n",
              "      <td>12.0</td>\n",
              "      <td>time:each</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>1.08626</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3345</th>\n",
              "      <td>NaN</td>\n",
              "      <td>13.0</td>\n",
              "      <td>time:each</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>1.08923</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3602</th>\n",
              "      <td>NaN</td>\n",
              "      <td>14.0</td>\n",
              "      <td>time:each</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>1.09282</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3859</th>\n",
              "      <td>NaN</td>\n",
              "      <td>15.0</td>\n",
              "      <td>time:each</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>1.10109</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4116</th>\n",
              "      <td>NaN</td>\n",
              "      <td>16.0</td>\n",
              "      <td>time:each</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>1.10515</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4373</th>\n",
              "      <td>NaN</td>\n",
              "      <td>17.0</td>\n",
              "      <td>time:each</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>1.10759</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4630</th>\n",
              "      <td>NaN</td>\n",
              "      <td>18.0</td>\n",
              "      <td>time:each</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>1.10776</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4887</th>\n",
              "      <td>NaN</td>\n",
              "      <td>19.0</td>\n",
              "      <td>time:each</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>1.10951</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "      i_batch  i_epoch        key      label    value\n",
              "5         NaN      0.0  time:each  per_epoch        0\n",
              "262       NaN      1.0  time:each  per_epoch   1.0845\n",
              "519       NaN      2.0  time:each  per_epoch  1.07928\n",
              "776       NaN      3.0  time:each  per_epoch  1.07654\n",
              "1033      NaN      4.0  time:each  per_epoch   1.0766\n",
              "1290      NaN      5.0  time:each  per_epoch  1.07655\n",
              "1547      NaN      6.0  time:each  per_epoch  1.07822\n",
              "1804      NaN      7.0  time:each  per_epoch  1.07816\n",
              "2061      NaN      8.0  time:each  per_epoch  1.07831\n",
              "2317      NaN      9.0  time:each  per_epoch  1.07786\n",
              "2574      NaN     10.0  time:each  per_epoch  1.08191\n",
              "2831      NaN     11.0  time:each  per_epoch  1.08398\n",
              "3088      NaN     12.0  time:each  per_epoch  1.08626\n",
              "3345      NaN     13.0  time:each  per_epoch  1.08923\n",
              "3602      NaN     14.0  time:each  per_epoch  1.09282\n",
              "3859      NaN     15.0  time:each  per_epoch  1.10109\n",
              "4116      NaN     16.0  time:each  per_epoch  1.10515\n",
              "4373      NaN     17.0  time:each  per_epoch  1.10759\n",
              "4630      NaN     18.0  time:each  per_epoch  1.10776\n",
              "4887      NaN     19.0  time:each  per_epoch  1.10951"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "per_epoch loss:mae\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>i_batch</th>\n",
              "      <th>i_epoch</th>\n",
              "      <th>key</th>\n",
              "      <th>label</th>\n",
              "      <th>value</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>246</th>\n",
              "      <td>NaN</td>\n",
              "      <td>0.0</td>\n",
              "      <td>loss:mae</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.489335</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>503</th>\n",
              "      <td>NaN</td>\n",
              "      <td>1.0</td>\n",
              "      <td>loss:mae</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.102269</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>760</th>\n",
              "      <td>NaN</td>\n",
              "      <td>2.0</td>\n",
              "      <td>loss:mae</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.0301363</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1017</th>\n",
              "      <td>NaN</td>\n",
              "      <td>3.0</td>\n",
              "      <td>loss:mae</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.0157085</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1274</th>\n",
              "      <td>NaN</td>\n",
              "      <td>4.0</td>\n",
              "      <td>loss:mae</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.0111635</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1531</th>\n",
              "      <td>NaN</td>\n",
              "      <td>5.0</td>\n",
              "      <td>loss:mae</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.00795995</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1788</th>\n",
              "      <td>NaN</td>\n",
              "      <td>6.0</td>\n",
              "      <td>loss:mae</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.0057916</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2045</th>\n",
              "      <td>NaN</td>\n",
              "      <td>7.0</td>\n",
              "      <td>loss:mae</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.00463259</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2302</th>\n",
              "      <td>NaN</td>\n",
              "      <td>8.0</td>\n",
              "      <td>loss:mae</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.00408322</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2558</th>\n",
              "      <td>NaN</td>\n",
              "      <td>9.0</td>\n",
              "      <td>loss:mae</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.00377736</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2815</th>\n",
              "      <td>NaN</td>\n",
              "      <td>10.0</td>\n",
              "      <td>loss:mae</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.00351545</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3072</th>\n",
              "      <td>NaN</td>\n",
              "      <td>11.0</td>\n",
              "      <td>loss:mae</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.00335075</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3329</th>\n",
              "      <td>NaN</td>\n",
              "      <td>12.0</td>\n",
              "      <td>loss:mae</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.00313241</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3586</th>\n",
              "      <td>NaN</td>\n",
              "      <td>13.0</td>\n",
              "      <td>loss:mae</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.00294693</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3843</th>\n",
              "      <td>NaN</td>\n",
              "      <td>14.0</td>\n",
              "      <td>loss:mae</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.00276498</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4100</th>\n",
              "      <td>NaN</td>\n",
              "      <td>15.0</td>\n",
              "      <td>loss:mae</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.00260081</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4357</th>\n",
              "      <td>NaN</td>\n",
              "      <td>16.0</td>\n",
              "      <td>loss:mae</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.0024506</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4614</th>\n",
              "      <td>NaN</td>\n",
              "      <td>17.0</td>\n",
              "      <td>loss:mae</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.00231502</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4871</th>\n",
              "      <td>NaN</td>\n",
              "      <td>18.0</td>\n",
              "      <td>loss:mae</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.00217059</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5128</th>\n",
              "      <td>NaN</td>\n",
              "      <td>19.0</td>\n",
              "      <td>loss:mae</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.00204612</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "      i_batch  i_epoch       key      label       value\n",
              "246       NaN      0.0  loss:mae  per_epoch    0.489335\n",
              "503       NaN      1.0  loss:mae  per_epoch    0.102269\n",
              "760       NaN      2.0  loss:mae  per_epoch   0.0301363\n",
              "1017      NaN      3.0  loss:mae  per_epoch   0.0157085\n",
              "1274      NaN      4.0  loss:mae  per_epoch   0.0111635\n",
              "1531      NaN      5.0  loss:mae  per_epoch  0.00795995\n",
              "1788      NaN      6.0  loss:mae  per_epoch   0.0057916\n",
              "2045      NaN      7.0  loss:mae  per_epoch  0.00463259\n",
              "2302      NaN      8.0  loss:mae  per_epoch  0.00408322\n",
              "2558      NaN      9.0  loss:mae  per_epoch  0.00377736\n",
              "2815      NaN     10.0  loss:mae  per_epoch  0.00351545\n",
              "3072      NaN     11.0  loss:mae  per_epoch  0.00335075\n",
              "3329      NaN     12.0  loss:mae  per_epoch  0.00313241\n",
              "3586      NaN     13.0  loss:mae  per_epoch  0.00294693\n",
              "3843      NaN     14.0  loss:mae  per_epoch  0.00276498\n",
              "4100      NaN     15.0  loss:mae  per_epoch  0.00260081\n",
              "4357      NaN     16.0  loss:mae  per_epoch   0.0024506\n",
              "4614      NaN     17.0  loss:mae  per_epoch  0.00231502\n",
              "4871      NaN     18.0  loss:mae  per_epoch  0.00217059\n",
              "5128      NaN     19.0  loss:mae  per_epoch  0.00204612"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "per_epoch loss:mse\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>i_batch</th>\n",
              "      <th>i_epoch</th>\n",
              "      <th>key</th>\n",
              "      <th>label</th>\n",
              "      <th>value</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>247</th>\n",
              "      <td>NaN</td>\n",
              "      <td>0.0</td>\n",
              "      <td>loss:mse</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.49112</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>504</th>\n",
              "      <td>NaN</td>\n",
              "      <td>1.0</td>\n",
              "      <td>loss:mse</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.0182449</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>761</th>\n",
              "      <td>NaN</td>\n",
              "      <td>2.0</td>\n",
              "      <td>loss:mse</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.00157353</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1018</th>\n",
              "      <td>NaN</td>\n",
              "      <td>3.0</td>\n",
              "      <td>loss:mse</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.000427152</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1275</th>\n",
              "      <td>NaN</td>\n",
              "      <td>4.0</td>\n",
              "      <td>loss:mse</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.000203408</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1532</th>\n",
              "      <td>NaN</td>\n",
              "      <td>5.0</td>\n",
              "      <td>loss:mse</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.000107406</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1789</th>\n",
              "      <td>NaN</td>\n",
              "      <td>6.0</td>\n",
              "      <td>loss:mse</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>6.13753e-05</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2046</th>\n",
              "      <td>NaN</td>\n",
              "      <td>7.0</td>\n",
              "      <td>loss:mse</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>4.11486e-05</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2303</th>\n",
              "      <td>NaN</td>\n",
              "      <td>8.0</td>\n",
              "      <td>loss:mse</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>3.11334e-05</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2559</th>\n",
              "      <td>NaN</td>\n",
              "      <td>9.0</td>\n",
              "      <td>loss:mse</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>2.59804e-05</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2816</th>\n",
              "      <td>NaN</td>\n",
              "      <td>10.0</td>\n",
              "      <td>loss:mse</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>2.2059e-05</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3073</th>\n",
              "      <td>NaN</td>\n",
              "      <td>11.0</td>\n",
              "      <td>loss:mse</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>1.96062e-05</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3330</th>\n",
              "      <td>NaN</td>\n",
              "      <td>12.0</td>\n",
              "      <td>loss:mse</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>1.69966e-05</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3587</th>\n",
              "      <td>NaN</td>\n",
              "      <td>13.0</td>\n",
              "      <td>loss:mse</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>1.5123e-05</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3844</th>\n",
              "      <td>NaN</td>\n",
              "      <td>14.0</td>\n",
              "      <td>loss:mse</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>1.34165e-05</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4101</th>\n",
              "      <td>NaN</td>\n",
              "      <td>15.0</td>\n",
              "      <td>loss:mse</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>1.2027e-05</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4358</th>\n",
              "      <td>NaN</td>\n",
              "      <td>16.0</td>\n",
              "      <td>loss:mse</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>1.06767e-05</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4615</th>\n",
              "      <td>NaN</td>\n",
              "      <td>17.0</td>\n",
              "      <td>loss:mse</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>9.73003e-06</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4872</th>\n",
              "      <td>NaN</td>\n",
              "      <td>18.0</td>\n",
              "      <td>loss:mse</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>8.78053e-06</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5129</th>\n",
              "      <td>NaN</td>\n",
              "      <td>19.0</td>\n",
              "      <td>loss:mse</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>7.98122e-06</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "      i_batch  i_epoch       key      label        value\n",
              "247       NaN      0.0  loss:mse  per_epoch      0.49112\n",
              "504       NaN      1.0  loss:mse  per_epoch    0.0182449\n",
              "761       NaN      2.0  loss:mse  per_epoch   0.00157353\n",
              "1018      NaN      3.0  loss:mse  per_epoch  0.000427152\n",
              "1275      NaN      4.0  loss:mse  per_epoch  0.000203408\n",
              "1532      NaN      5.0  loss:mse  per_epoch  0.000107406\n",
              "1789      NaN      6.0  loss:mse  per_epoch  6.13753e-05\n",
              "2046      NaN      7.0  loss:mse  per_epoch  4.11486e-05\n",
              "2303      NaN      8.0  loss:mse  per_epoch  3.11334e-05\n",
              "2559      NaN      9.0  loss:mse  per_epoch  2.59804e-05\n",
              "2816      NaN     10.0  loss:mse  per_epoch   2.2059e-05\n",
              "3073      NaN     11.0  loss:mse  per_epoch  1.96062e-05\n",
              "3330      NaN     12.0  loss:mse  per_epoch  1.69966e-05\n",
              "3587      NaN     13.0  loss:mse  per_epoch   1.5123e-05\n",
              "3844      NaN     14.0  loss:mse  per_epoch  1.34165e-05\n",
              "4101      NaN     15.0  loss:mse  per_epoch   1.2027e-05\n",
              "4358      NaN     16.0  loss:mse  per_epoch  1.06767e-05\n",
              "4615      NaN     17.0  loss:mse  per_epoch  9.73003e-06\n",
              "4872      NaN     18.0  loss:mse  per_epoch  8.78053e-06\n",
              "5129      NaN     19.0  loss:mse  per_epoch  7.98122e-06"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "per_epoch metric:1\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>i_batch</th>\n",
              "      <th>i_epoch</th>\n",
              "      <th>key</th>\n",
              "      <th>label</th>\n",
              "      <th>value</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>248</th>\n",
              "      <td>NaN</td>\n",
              "      <td>0.0</td>\n",
              "      <td>metric:1</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.980455</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>253</th>\n",
              "      <td>NaN</td>\n",
              "      <td>0.0</td>\n",
              "      <td>metric:1</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.864738</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>505</th>\n",
              "      <td>NaN</td>\n",
              "      <td>1.0</td>\n",
              "      <td>metric:1</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.120514</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>510</th>\n",
              "      <td>NaN</td>\n",
              "      <td>1.0</td>\n",
              "      <td>metric:1</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.675478</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>762</th>\n",
              "      <td>NaN</td>\n",
              "      <td>2.0</td>\n",
              "      <td>metric:1</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.0317098</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>767</th>\n",
              "      <td>NaN</td>\n",
              "      <td>2.0</td>\n",
              "      <td>metric:1</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.55632</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1019</th>\n",
              "      <td>NaN</td>\n",
              "      <td>3.0</td>\n",
              "      <td>metric:1</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.0161356</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1024</th>\n",
              "      <td>NaN</td>\n",
              "      <td>3.0</td>\n",
              "      <td>metric:1</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.463872</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1276</th>\n",
              "      <td>NaN</td>\n",
              "      <td>4.0</td>\n",
              "      <td>metric:1</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.0113669</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1281</th>\n",
              "      <td>NaN</td>\n",
              "      <td>4.0</td>\n",
              "      <td>metric:1</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.415771</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1533</th>\n",
              "      <td>NaN</td>\n",
              "      <td>5.0</td>\n",
              "      <td>metric:1</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.00806736</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1538</th>\n",
              "      <td>NaN</td>\n",
              "      <td>5.0</td>\n",
              "      <td>metric:1</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.405454</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1790</th>\n",
              "      <td>NaN</td>\n",
              "      <td>6.0</td>\n",
              "      <td>metric:1</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.00585298</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1795</th>\n",
              "      <td>NaN</td>\n",
              "      <td>6.0</td>\n",
              "      <td>metric:1</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.378996</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2047</th>\n",
              "      <td>NaN</td>\n",
              "      <td>7.0</td>\n",
              "      <td>metric:1</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.00467374</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2052</th>\n",
              "      <td>NaN</td>\n",
              "      <td>7.0</td>\n",
              "      <td>metric:1</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.361285</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2304</th>\n",
              "      <td>NaN</td>\n",
              "      <td>8.0</td>\n",
              "      <td>metric:1</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.00411435</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2309</th>\n",
              "      <td>NaN</td>\n",
              "      <td>8.0</td>\n",
              "      <td>metric:1</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.362011</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2560</th>\n",
              "      <td>NaN</td>\n",
              "      <td>9.0</td>\n",
              "      <td>metric:1</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.00380334</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2565</th>\n",
              "      <td>NaN</td>\n",
              "      <td>9.0</td>\n",
              "      <td>metric:1</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.360873</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2817</th>\n",
              "      <td>NaN</td>\n",
              "      <td>10.0</td>\n",
              "      <td>metric:1</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.00353751</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2822</th>\n",
              "      <td>NaN</td>\n",
              "      <td>10.0</td>\n",
              "      <td>metric:1</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.347469</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3074</th>\n",
              "      <td>NaN</td>\n",
              "      <td>11.0</td>\n",
              "      <td>metric:1</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.00337036</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3079</th>\n",
              "      <td>NaN</td>\n",
              "      <td>11.0</td>\n",
              "      <td>metric:1</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.341871</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3331</th>\n",
              "      <td>NaN</td>\n",
              "      <td>12.0</td>\n",
              "      <td>metric:1</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.00314941</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3336</th>\n",
              "      <td>NaN</td>\n",
              "      <td>12.0</td>\n",
              "      <td>metric:1</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.333861</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3588</th>\n",
              "      <td>NaN</td>\n",
              "      <td>13.0</td>\n",
              "      <td>metric:1</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.00296206</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3593</th>\n",
              "      <td>NaN</td>\n",
              "      <td>13.0</td>\n",
              "      <td>metric:1</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.331496</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3845</th>\n",
              "      <td>NaN</td>\n",
              "      <td>14.0</td>\n",
              "      <td>metric:1</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.0027784</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3850</th>\n",
              "      <td>NaN</td>\n",
              "      <td>14.0</td>\n",
              "      <td>metric:1</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.329377</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4102</th>\n",
              "      <td>NaN</td>\n",
              "      <td>15.0</td>\n",
              "      <td>metric:1</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.00261283</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4107</th>\n",
              "      <td>NaN</td>\n",
              "      <td>15.0</td>\n",
              "      <td>metric:1</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.326495</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4359</th>\n",
              "      <td>NaN</td>\n",
              "      <td>16.0</td>\n",
              "      <td>metric:1</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.00246127</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4364</th>\n",
              "      <td>NaN</td>\n",
              "      <td>16.0</td>\n",
              "      <td>metric:1</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.325285</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4616</th>\n",
              "      <td>NaN</td>\n",
              "      <td>17.0</td>\n",
              "      <td>metric:1</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.00232475</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4621</th>\n",
              "      <td>NaN</td>\n",
              "      <td>17.0</td>\n",
              "      <td>metric:1</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.319591</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4873</th>\n",
              "      <td>NaN</td>\n",
              "      <td>18.0</td>\n",
              "      <td>metric:1</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.00217938</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4878</th>\n",
              "      <td>NaN</td>\n",
              "      <td>18.0</td>\n",
              "      <td>metric:1</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.315575</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5130</th>\n",
              "      <td>NaN</td>\n",
              "      <td>19.0</td>\n",
              "      <td>metric:1</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.0020541</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5135</th>\n",
              "      <td>NaN</td>\n",
              "      <td>19.0</td>\n",
              "      <td>metric:1</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.3117</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "      i_batch  i_epoch       key      label       value\n",
              "248       NaN      0.0  metric:1  per_epoch    0.980455\n",
              "253       NaN      0.0  metric:1  per_epoch    0.864738\n",
              "505       NaN      1.0  metric:1  per_epoch    0.120514\n",
              "510       NaN      1.0  metric:1  per_epoch    0.675478\n",
              "762       NaN      2.0  metric:1  per_epoch   0.0317098\n",
              "767       NaN      2.0  metric:1  per_epoch     0.55632\n",
              "1019      NaN      3.0  metric:1  per_epoch   0.0161356\n",
              "1024      NaN      3.0  metric:1  per_epoch    0.463872\n",
              "1276      NaN      4.0  metric:1  per_epoch   0.0113669\n",
              "1281      NaN      4.0  metric:1  per_epoch    0.415771\n",
              "1533      NaN      5.0  metric:1  per_epoch  0.00806736\n",
              "1538      NaN      5.0  metric:1  per_epoch    0.405454\n",
              "1790      NaN      6.0  metric:1  per_epoch  0.00585298\n",
              "1795      NaN      6.0  metric:1  per_epoch    0.378996\n",
              "2047      NaN      7.0  metric:1  per_epoch  0.00467374\n",
              "2052      NaN      7.0  metric:1  per_epoch    0.361285\n",
              "2304      NaN      8.0  metric:1  per_epoch  0.00411435\n",
              "2309      NaN      8.0  metric:1  per_epoch    0.362011\n",
              "2560      NaN      9.0  metric:1  per_epoch  0.00380334\n",
              "2565      NaN      9.0  metric:1  per_epoch    0.360873\n",
              "2817      NaN     10.0  metric:1  per_epoch  0.00353751\n",
              "2822      NaN     10.0  metric:1  per_epoch    0.347469\n",
              "3074      NaN     11.0  metric:1  per_epoch  0.00337036\n",
              "3079      NaN     11.0  metric:1  per_epoch    0.341871\n",
              "3331      NaN     12.0  metric:1  per_epoch  0.00314941\n",
              "3336      NaN     12.0  metric:1  per_epoch    0.333861\n",
              "3588      NaN     13.0  metric:1  per_epoch  0.00296206\n",
              "3593      NaN     13.0  metric:1  per_epoch    0.331496\n",
              "3845      NaN     14.0  metric:1  per_epoch   0.0027784\n",
              "3850      NaN     14.0  metric:1  per_epoch    0.329377\n",
              "4102      NaN     15.0  metric:1  per_epoch  0.00261283\n",
              "4107      NaN     15.0  metric:1  per_epoch    0.326495\n",
              "4359      NaN     16.0  metric:1  per_epoch  0.00246127\n",
              "4364      NaN     16.0  metric:1  per_epoch    0.325285\n",
              "4616      NaN     17.0  metric:1  per_epoch  0.00232475\n",
              "4621      NaN     17.0  metric:1  per_epoch    0.319591\n",
              "4873      NaN     18.0  metric:1  per_epoch  0.00217938\n",
              "4878      NaN     18.0  metric:1  per_epoch    0.315575\n",
              "5130      NaN     19.0  metric:1  per_epoch   0.0020541\n",
              "5135      NaN     19.0  metric:1  per_epoch      0.3117"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "per_epoch metric:2\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>i_batch</th>\n",
              "      <th>i_epoch</th>\n",
              "      <th>key</th>\n",
              "      <th>label</th>\n",
              "      <th>value</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>249</th>\n",
              "      <td>NaN</td>\n",
              "      <td>0.0</td>\n",
              "      <td>metric:2</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>-0.00178569</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>254</th>\n",
              "      <td>NaN</td>\n",
              "      <td>0.0</td>\n",
              "      <td>metric:2</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.0750187</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>506</th>\n",
              "      <td>NaN</td>\n",
              "      <td>1.0</td>\n",
              "      <td>metric:2</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.0840241</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>511</th>\n",
              "      <td>NaN</td>\n",
              "      <td>1.0</td>\n",
              "      <td>metric:2</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>-0.0306056</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>763</th>\n",
              "      <td>NaN</td>\n",
              "      <td>2.0</td>\n",
              "      <td>metric:2</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.0285627</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>768</th>\n",
              "      <td>NaN</td>\n",
              "      <td>2.0</td>\n",
              "      <td>metric:2</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>-0.00848353</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1020</th>\n",
              "      <td>NaN</td>\n",
              "      <td>3.0</td>\n",
              "      <td>metric:2</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.0152813</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1025</th>\n",
              "      <td>NaN</td>\n",
              "      <td>3.0</td>\n",
              "      <td>metric:2</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.0118707</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1277</th>\n",
              "      <td>NaN</td>\n",
              "      <td>4.0</td>\n",
              "      <td>metric:2</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.01096</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1282</th>\n",
              "      <td>NaN</td>\n",
              "      <td>4.0</td>\n",
              "      <td>metric:2</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.0187656</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1534</th>\n",
              "      <td>NaN</td>\n",
              "      <td>5.0</td>\n",
              "      <td>metric:2</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.00785254</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1539</th>\n",
              "      <td>NaN</td>\n",
              "      <td>5.0</td>\n",
              "      <td>metric:2</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.0201903</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1791</th>\n",
              "      <td>NaN</td>\n",
              "      <td>6.0</td>\n",
              "      <td>metric:2</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.00573023</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1796</th>\n",
              "      <td>NaN</td>\n",
              "      <td>6.0</td>\n",
              "      <td>metric:2</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.0228509</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2048</th>\n",
              "      <td>NaN</td>\n",
              "      <td>7.0</td>\n",
              "      <td>metric:2</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.00459144</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2053</th>\n",
              "      <td>NaN</td>\n",
              "      <td>7.0</td>\n",
              "      <td>metric:2</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.0251761</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2305</th>\n",
              "      <td>NaN</td>\n",
              "      <td>8.0</td>\n",
              "      <td>metric:2</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.00405209</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2310</th>\n",
              "      <td>NaN</td>\n",
              "      <td>8.0</td>\n",
              "      <td>metric:2</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.0254832</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2561</th>\n",
              "      <td>NaN</td>\n",
              "      <td>9.0</td>\n",
              "      <td>metric:2</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.00375138</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2566</th>\n",
              "      <td>NaN</td>\n",
              "      <td>9.0</td>\n",
              "      <td>metric:2</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.0275335</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2818</th>\n",
              "      <td>NaN</td>\n",
              "      <td>10.0</td>\n",
              "      <td>metric:2</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.00349339</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2823</th>\n",
              "      <td>NaN</td>\n",
              "      <td>10.0</td>\n",
              "      <td>metric:2</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.0283843</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3075</th>\n",
              "      <td>NaN</td>\n",
              "      <td>11.0</td>\n",
              "      <td>metric:2</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.00333115</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3080</th>\n",
              "      <td>NaN</td>\n",
              "      <td>11.0</td>\n",
              "      <td>metric:2</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.0293948</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3332</th>\n",
              "      <td>NaN</td>\n",
              "      <td>12.0</td>\n",
              "      <td>metric:2</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.00311541</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3337</th>\n",
              "      <td>NaN</td>\n",
              "      <td>12.0</td>\n",
              "      <td>metric:2</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.030268</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3589</th>\n",
              "      <td>NaN</td>\n",
              "      <td>13.0</td>\n",
              "      <td>metric:2</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.00293181</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3594</th>\n",
              "      <td>NaN</td>\n",
              "      <td>13.0</td>\n",
              "      <td>metric:2</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.0320252</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3846</th>\n",
              "      <td>NaN</td>\n",
              "      <td>14.0</td>\n",
              "      <td>metric:2</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.00275157</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3851</th>\n",
              "      <td>NaN</td>\n",
              "      <td>14.0</td>\n",
              "      <td>metric:2</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.0321252</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4103</th>\n",
              "      <td>NaN</td>\n",
              "      <td>15.0</td>\n",
              "      <td>metric:2</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.00258878</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4108</th>\n",
              "      <td>NaN</td>\n",
              "      <td>15.0</td>\n",
              "      <td>metric:2</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.0307765</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4360</th>\n",
              "      <td>NaN</td>\n",
              "      <td>16.0</td>\n",
              "      <td>metric:2</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.00243992</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4365</th>\n",
              "      <td>NaN</td>\n",
              "      <td>16.0</td>\n",
              "      <td>metric:2</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.0323822</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4617</th>\n",
              "      <td>NaN</td>\n",
              "      <td>17.0</td>\n",
              "      <td>metric:2</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.00230529</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4622</th>\n",
              "      <td>NaN</td>\n",
              "      <td>17.0</td>\n",
              "      <td>metric:2</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.0329631</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4874</th>\n",
              "      <td>NaN</td>\n",
              "      <td>18.0</td>\n",
              "      <td>metric:2</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.00216181</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4879</th>\n",
              "      <td>NaN</td>\n",
              "      <td>18.0</td>\n",
              "      <td>metric:2</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.0338162</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5131</th>\n",
              "      <td>NaN</td>\n",
              "      <td>19.0</td>\n",
              "      <td>metric:2</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.00203814</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5136</th>\n",
              "      <td>NaN</td>\n",
              "      <td>19.0</td>\n",
              "      <td>metric:2</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.0338856</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "      i_batch  i_epoch       key      label       value\n",
              "249       NaN      0.0  metric:2  per_epoch -0.00178569\n",
              "254       NaN      0.0  metric:2  per_epoch   0.0750187\n",
              "506       NaN      1.0  metric:2  per_epoch   0.0840241\n",
              "511       NaN      1.0  metric:2  per_epoch  -0.0306056\n",
              "763       NaN      2.0  metric:2  per_epoch   0.0285627\n",
              "768       NaN      2.0  metric:2  per_epoch -0.00848353\n",
              "1020      NaN      3.0  metric:2  per_epoch   0.0152813\n",
              "1025      NaN      3.0  metric:2  per_epoch   0.0118707\n",
              "1277      NaN      4.0  metric:2  per_epoch     0.01096\n",
              "1282      NaN      4.0  metric:2  per_epoch   0.0187656\n",
              "1534      NaN      5.0  metric:2  per_epoch  0.00785254\n",
              "1539      NaN      5.0  metric:2  per_epoch   0.0201903\n",
              "1791      NaN      6.0  metric:2  per_epoch  0.00573023\n",
              "1796      NaN      6.0  metric:2  per_epoch   0.0228509\n",
              "2048      NaN      7.0  metric:2  per_epoch  0.00459144\n",
              "2053      NaN      7.0  metric:2  per_epoch   0.0251761\n",
              "2305      NaN      8.0  metric:2  per_epoch  0.00405209\n",
              "2310      NaN      8.0  metric:2  per_epoch   0.0254832\n",
              "2561      NaN      9.0  metric:2  per_epoch  0.00375138\n",
              "2566      NaN      9.0  metric:2  per_epoch   0.0275335\n",
              "2818      NaN     10.0  metric:2  per_epoch  0.00349339\n",
              "2823      NaN     10.0  metric:2  per_epoch   0.0283843\n",
              "3075      NaN     11.0  metric:2  per_epoch  0.00333115\n",
              "3080      NaN     11.0  metric:2  per_epoch   0.0293948\n",
              "3332      NaN     12.0  metric:2  per_epoch  0.00311541\n",
              "3337      NaN     12.0  metric:2  per_epoch    0.030268\n",
              "3589      NaN     13.0  metric:2  per_epoch  0.00293181\n",
              "3594      NaN     13.0  metric:2  per_epoch   0.0320252\n",
              "3846      NaN     14.0  metric:2  per_epoch  0.00275157\n",
              "3851      NaN     14.0  metric:2  per_epoch   0.0321252\n",
              "4103      NaN     15.0  metric:2  per_epoch  0.00258878\n",
              "4108      NaN     15.0  metric:2  per_epoch   0.0307765\n",
              "4360      NaN     16.0  metric:2  per_epoch  0.00243992\n",
              "4365      NaN     16.0  metric:2  per_epoch   0.0323822\n",
              "4617      NaN     17.0  metric:2  per_epoch  0.00230529\n",
              "4622      NaN     17.0  metric:2  per_epoch   0.0329631\n",
              "4874      NaN     18.0  metric:2  per_epoch  0.00216181\n",
              "4879      NaN     18.0  metric:2  per_epoch   0.0338162\n",
              "5131      NaN     19.0  metric:2  per_epoch  0.00203814\n",
              "5136      NaN     19.0  metric:2  per_epoch   0.0338856"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "per_epoch gpu-mem\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>i_batch</th>\n",
              "      <th>i_epoch</th>\n",
              "      <th>key</th>\n",
              "      <th>label</th>\n",
              "      <th>value</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>250</th>\n",
              "      <td>NaN</td>\n",
              "      <td>0.0</td>\n",
              "      <td>gpu-mem</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>203423744</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>507</th>\n",
              "      <td>NaN</td>\n",
              "      <td>1.0</td>\n",
              "      <td>gpu-mem</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>874512384</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>764</th>\n",
              "      <td>NaN</td>\n",
              "      <td>2.0</td>\n",
              "      <td>gpu-mem</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>874512384</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1021</th>\n",
              "      <td>NaN</td>\n",
              "      <td>3.0</td>\n",
              "      <td>gpu-mem</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>874512384</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1278</th>\n",
              "      <td>NaN</td>\n",
              "      <td>4.0</td>\n",
              "      <td>gpu-mem</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>874512384</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1535</th>\n",
              "      <td>NaN</td>\n",
              "      <td>5.0</td>\n",
              "      <td>gpu-mem</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>874512384</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1792</th>\n",
              "      <td>NaN</td>\n",
              "      <td>6.0</td>\n",
              "      <td>gpu-mem</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>874512384</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2049</th>\n",
              "      <td>NaN</td>\n",
              "      <td>7.0</td>\n",
              "      <td>gpu-mem</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>874512384</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2306</th>\n",
              "      <td>NaN</td>\n",
              "      <td>8.0</td>\n",
              "      <td>gpu-mem</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>874512384</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2562</th>\n",
              "      <td>NaN</td>\n",
              "      <td>9.0</td>\n",
              "      <td>gpu-mem</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>874512384</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2819</th>\n",
              "      <td>NaN</td>\n",
              "      <td>10.0</td>\n",
              "      <td>gpu-mem</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>874512384</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3076</th>\n",
              "      <td>NaN</td>\n",
              "      <td>11.0</td>\n",
              "      <td>gpu-mem</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>874512384</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3333</th>\n",
              "      <td>NaN</td>\n",
              "      <td>12.0</td>\n",
              "      <td>gpu-mem</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>874512384</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3590</th>\n",
              "      <td>NaN</td>\n",
              "      <td>13.0</td>\n",
              "      <td>gpu-mem</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>874512384</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3847</th>\n",
              "      <td>NaN</td>\n",
              "      <td>14.0</td>\n",
              "      <td>gpu-mem</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>874512384</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4104</th>\n",
              "      <td>NaN</td>\n",
              "      <td>15.0</td>\n",
              "      <td>gpu-mem</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>874512384</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4361</th>\n",
              "      <td>NaN</td>\n",
              "      <td>16.0</td>\n",
              "      <td>gpu-mem</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>874512384</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4618</th>\n",
              "      <td>NaN</td>\n",
              "      <td>17.0</td>\n",
              "      <td>gpu-mem</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>874512384</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4875</th>\n",
              "      <td>NaN</td>\n",
              "      <td>18.0</td>\n",
              "      <td>gpu-mem</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>874512384</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5132</th>\n",
              "      <td>NaN</td>\n",
              "      <td>19.0</td>\n",
              "      <td>gpu-mem</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>874512384</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "      i_batch  i_epoch      key      label      value\n",
              "250       NaN      0.0  gpu-mem  per_epoch  203423744\n",
              "507       NaN      1.0  gpu-mem  per_epoch  874512384\n",
              "764       NaN      2.0  gpu-mem  per_epoch  874512384\n",
              "1021      NaN      3.0  gpu-mem  per_epoch  874512384\n",
              "1278      NaN      4.0  gpu-mem  per_epoch  874512384\n",
              "1535      NaN      5.0  gpu-mem  per_epoch  874512384\n",
              "1792      NaN      6.0  gpu-mem  per_epoch  874512384\n",
              "2049      NaN      7.0  gpu-mem  per_epoch  874512384\n",
              "2306      NaN      8.0  gpu-mem  per_epoch  874512384\n",
              "2562      NaN      9.0  gpu-mem  per_epoch  874512384\n",
              "2819      NaN     10.0  gpu-mem  per_epoch  874512384\n",
              "3076      NaN     11.0  gpu-mem  per_epoch  874512384\n",
              "3333      NaN     12.0  gpu-mem  per_epoch  874512384\n",
              "3590      NaN     13.0  gpu-mem  per_epoch  874512384\n",
              "3847      NaN     14.0  gpu-mem  per_epoch  874512384\n",
              "4104      NaN     15.0  gpu-mem  per_epoch  874512384\n",
              "4361      NaN     16.0  gpu-mem  per_epoch  874512384\n",
              "4618      NaN     17.0  gpu-mem  per_epoch  874512384\n",
              "4875      NaN     18.0  gpu-mem  per_epoch  874512384\n",
              "5132      NaN     19.0  gpu-mem  per_epoch  874512384"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "per_epoch loss:val_mae\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>i_batch</th>\n",
              "      <th>i_epoch</th>\n",
              "      <th>key</th>\n",
              "      <th>label</th>\n",
              "      <th>value</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>251</th>\n",
              "      <td>NaN</td>\n",
              "      <td>0.0</td>\n",
              "      <td>loss:val_mae</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.469878</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>508</th>\n",
              "      <td>NaN</td>\n",
              "      <td>1.0</td>\n",
              "      <td>loss:val_mae</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.322436</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>765</th>\n",
              "      <td>NaN</td>\n",
              "      <td>2.0</td>\n",
              "      <td>loss:val_mae</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.273918</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1022</th>\n",
              "      <td>NaN</td>\n",
              "      <td>3.0</td>\n",
              "      <td>loss:val_mae</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.237871</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1279</th>\n",
              "      <td>NaN</td>\n",
              "      <td>4.0</td>\n",
              "      <td>loss:val_mae</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.217268</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1536</th>\n",
              "      <td>NaN</td>\n",
              "      <td>5.0</td>\n",
              "      <td>loss:val_mae</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.212822</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1793</th>\n",
              "      <td>NaN</td>\n",
              "      <td>6.0</td>\n",
              "      <td>loss:val_mae</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.200923</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2050</th>\n",
              "      <td>NaN</td>\n",
              "      <td>7.0</td>\n",
              "      <td>loss:val_mae</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.19323</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2307</th>\n",
              "      <td>NaN</td>\n",
              "      <td>8.0</td>\n",
              "      <td>loss:val_mae</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.193747</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2563</th>\n",
              "      <td>NaN</td>\n",
              "      <td>9.0</td>\n",
              "      <td>loss:val_mae</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.194203</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2820</th>\n",
              "      <td>NaN</td>\n",
              "      <td>10.0</td>\n",
              "      <td>loss:val_mae</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.187926</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3077</th>\n",
              "      <td>NaN</td>\n",
              "      <td>11.0</td>\n",
              "      <td>loss:val_mae</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.185633</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3334</th>\n",
              "      <td>NaN</td>\n",
              "      <td>12.0</td>\n",
              "      <td>loss:val_mae</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.182064</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3591</th>\n",
              "      <td>NaN</td>\n",
              "      <td>13.0</td>\n",
              "      <td>loss:val_mae</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.18176</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3848</th>\n",
              "      <td>NaN</td>\n",
              "      <td>14.0</td>\n",
              "      <td>loss:val_mae</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.180751</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4105</th>\n",
              "      <td>NaN</td>\n",
              "      <td>15.0</td>\n",
              "      <td>loss:val_mae</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.178636</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4362</th>\n",
              "      <td>NaN</td>\n",
              "      <td>16.0</td>\n",
              "      <td>loss:val_mae</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.178833</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4619</th>\n",
              "      <td>NaN</td>\n",
              "      <td>17.0</td>\n",
              "      <td>loss:val_mae</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.176277</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4876</th>\n",
              "      <td>NaN</td>\n",
              "      <td>18.0</td>\n",
              "      <td>loss:val_mae</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.174696</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5133</th>\n",
              "      <td>NaN</td>\n",
              "      <td>19.0</td>\n",
              "      <td>loss:val_mae</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.172793</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "      i_batch  i_epoch           key      label     value\n",
              "251       NaN      0.0  loss:val_mae  per_epoch  0.469878\n",
              "508       NaN      1.0  loss:val_mae  per_epoch  0.322436\n",
              "765       NaN      2.0  loss:val_mae  per_epoch  0.273918\n",
              "1022      NaN      3.0  loss:val_mae  per_epoch  0.237871\n",
              "1279      NaN      4.0  loss:val_mae  per_epoch  0.217268\n",
              "1536      NaN      5.0  loss:val_mae  per_epoch  0.212822\n",
              "1793      NaN      6.0  loss:val_mae  per_epoch  0.200923\n",
              "2050      NaN      7.0  loss:val_mae  per_epoch   0.19323\n",
              "2307      NaN      8.0  loss:val_mae  per_epoch  0.193747\n",
              "2563      NaN      9.0  loss:val_mae  per_epoch  0.194203\n",
              "2820      NaN     10.0  loss:val_mae  per_epoch  0.187926\n",
              "3077      NaN     11.0  loss:val_mae  per_epoch  0.185633\n",
              "3334      NaN     12.0  loss:val_mae  per_epoch  0.182064\n",
              "3591      NaN     13.0  loss:val_mae  per_epoch   0.18176\n",
              "3848      NaN     14.0  loss:val_mae  per_epoch  0.180751\n",
              "4105      NaN     15.0  loss:val_mae  per_epoch  0.178636\n",
              "4362      NaN     16.0  loss:val_mae  per_epoch  0.178833\n",
              "4619      NaN     17.0  loss:val_mae  per_epoch  0.176277\n",
              "4876      NaN     18.0  loss:val_mae  per_epoch  0.174696\n",
              "5133      NaN     19.0  loss:val_mae  per_epoch  0.172793"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "per_epoch loss:val_mse\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>i_batch</th>\n",
              "      <th>i_epoch</th>\n",
              "      <th>key</th>\n",
              "      <th>label</th>\n",
              "      <th>value</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>252</th>\n",
              "      <td>NaN</td>\n",
              "      <td>0.0</td>\n",
              "      <td>loss:val_mse</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.394859</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>509</th>\n",
              "      <td>NaN</td>\n",
              "      <td>1.0</td>\n",
              "      <td>loss:val_mse</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.353042</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>766</th>\n",
              "      <td>NaN</td>\n",
              "      <td>2.0</td>\n",
              "      <td>loss:val_mse</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.282402</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1023</th>\n",
              "      <td>NaN</td>\n",
              "      <td>3.0</td>\n",
              "      <td>loss:val_mse</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.226</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1280</th>\n",
              "      <td>NaN</td>\n",
              "      <td>4.0</td>\n",
              "      <td>loss:val_mse</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.198503</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1537</th>\n",
              "      <td>NaN</td>\n",
              "      <td>5.0</td>\n",
              "      <td>loss:val_mse</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.192632</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1794</th>\n",
              "      <td>NaN</td>\n",
              "      <td>6.0</td>\n",
              "      <td>loss:val_mse</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.178073</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2051</th>\n",
              "      <td>NaN</td>\n",
              "      <td>7.0</td>\n",
              "      <td>loss:val_mse</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.168054</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2308</th>\n",
              "      <td>NaN</td>\n",
              "      <td>8.0</td>\n",
              "      <td>loss:val_mse</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.168264</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2564</th>\n",
              "      <td>NaN</td>\n",
              "      <td>9.0</td>\n",
              "      <td>loss:val_mse</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.16667</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2821</th>\n",
              "      <td>NaN</td>\n",
              "      <td>10.0</td>\n",
              "      <td>loss:val_mse</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.159542</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3078</th>\n",
              "      <td>NaN</td>\n",
              "      <td>11.0</td>\n",
              "      <td>loss:val_mse</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.156238</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3335</th>\n",
              "      <td>NaN</td>\n",
              "      <td>12.0</td>\n",
              "      <td>loss:val_mse</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.151796</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3592</th>\n",
              "      <td>NaN</td>\n",
              "      <td>13.0</td>\n",
              "      <td>loss:val_mse</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.149735</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3849</th>\n",
              "      <td>NaN</td>\n",
              "      <td>14.0</td>\n",
              "      <td>loss:val_mse</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.148626</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4106</th>\n",
              "      <td>NaN</td>\n",
              "      <td>15.0</td>\n",
              "      <td>loss:val_mse</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.147859</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4363</th>\n",
              "      <td>NaN</td>\n",
              "      <td>16.0</td>\n",
              "      <td>loss:val_mse</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.146451</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4620</th>\n",
              "      <td>NaN</td>\n",
              "      <td>17.0</td>\n",
              "      <td>loss:val_mse</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.143314</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4877</th>\n",
              "      <td>NaN</td>\n",
              "      <td>18.0</td>\n",
              "      <td>loss:val_mse</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.140879</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5134</th>\n",
              "      <td>NaN</td>\n",
              "      <td>19.0</td>\n",
              "      <td>loss:val_mse</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>0.138907</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "      i_batch  i_epoch           key      label     value\n",
              "252       NaN      0.0  loss:val_mse  per_epoch  0.394859\n",
              "509       NaN      1.0  loss:val_mse  per_epoch  0.353042\n",
              "766       NaN      2.0  loss:val_mse  per_epoch  0.282402\n",
              "1023      NaN      3.0  loss:val_mse  per_epoch     0.226\n",
              "1280      NaN      4.0  loss:val_mse  per_epoch  0.198503\n",
              "1537      NaN      5.0  loss:val_mse  per_epoch  0.192632\n",
              "1794      NaN      6.0  loss:val_mse  per_epoch  0.178073\n",
              "2051      NaN      7.0  loss:val_mse  per_epoch  0.168054\n",
              "2308      NaN      8.0  loss:val_mse  per_epoch  0.168264\n",
              "2564      NaN      9.0  loss:val_mse  per_epoch   0.16667\n",
              "2821      NaN     10.0  loss:val_mse  per_epoch  0.159542\n",
              "3078      NaN     11.0  loss:val_mse  per_epoch  0.156238\n",
              "3335      NaN     12.0  loss:val_mse  per_epoch  0.151796\n",
              "3592      NaN     13.0  loss:val_mse  per_epoch  0.149735\n",
              "3849      NaN     14.0  loss:val_mse  per_epoch  0.148626\n",
              "4106      NaN     15.0  loss:val_mse  per_epoch  0.147859\n",
              "4363      NaN     16.0  loss:val_mse  per_epoch  0.146451\n",
              "4620      NaN     17.0  loss:val_mse  per_epoch  0.143314\n",
              "4877      NaN     18.0  loss:val_mse  per_epoch  0.140879\n",
              "5134      NaN     19.0  loss:val_mse  per_epoch  0.138907"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "per_epoch model:net_state_dict\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>i_batch</th>\n",
              "      <th>i_epoch</th>\n",
              "      <th>key</th>\n",
              "      <th>label</th>\n",
              "      <th>value</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>255</th>\n",
              "      <td>NaN</td>\n",
              "      <td>0.0</td>\n",
              "      <td>model:net_state_dict</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>{'0.weight': [[tensor(0.5510, device='cuda:0')...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>512</th>\n",
              "      <td>NaN</td>\n",
              "      <td>1.0</td>\n",
              "      <td>model:net_state_dict</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>{'0.weight': [[tensor(0.5510, device='cuda:0')...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>769</th>\n",
              "      <td>NaN</td>\n",
              "      <td>2.0</td>\n",
              "      <td>model:net_state_dict</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>{'0.weight': [[tensor(0.5510, device='cuda:0')...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1026</th>\n",
              "      <td>NaN</td>\n",
              "      <td>3.0</td>\n",
              "      <td>model:net_state_dict</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>{'0.weight': [[tensor(0.5510, device='cuda:0')...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1283</th>\n",
              "      <td>NaN</td>\n",
              "      <td>4.0</td>\n",
              "      <td>model:net_state_dict</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>{'0.weight': [[tensor(0.5510, device='cuda:0')...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1540</th>\n",
              "      <td>NaN</td>\n",
              "      <td>5.0</td>\n",
              "      <td>model:net_state_dict</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>{'0.weight': [[tensor(0.5510, device='cuda:0')...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1797</th>\n",
              "      <td>NaN</td>\n",
              "      <td>6.0</td>\n",
              "      <td>model:net_state_dict</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>{'0.weight': [[tensor(0.5510, device='cuda:0')...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2054</th>\n",
              "      <td>NaN</td>\n",
              "      <td>7.0</td>\n",
              "      <td>model:net_state_dict</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>{'0.weight': [[tensor(0.5510, device='cuda:0')...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2311</th>\n",
              "      <td>NaN</td>\n",
              "      <td>8.0</td>\n",
              "      <td>model:net_state_dict</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>{'0.weight': [[tensor(0.5510, device='cuda:0')...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2567</th>\n",
              "      <td>NaN</td>\n",
              "      <td>9.0</td>\n",
              "      <td>model:net_state_dict</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>{'0.weight': [[tensor(0.5510, device='cuda:0')...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2824</th>\n",
              "      <td>NaN</td>\n",
              "      <td>10.0</td>\n",
              "      <td>model:net_state_dict</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>{'0.weight': [[tensor(0.5510, device='cuda:0')...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3081</th>\n",
              "      <td>NaN</td>\n",
              "      <td>11.0</td>\n",
              "      <td>model:net_state_dict</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>{'0.weight': [[tensor(0.5510, device='cuda:0')...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3338</th>\n",
              "      <td>NaN</td>\n",
              "      <td>12.0</td>\n",
              "      <td>model:net_state_dict</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>{'0.weight': [[tensor(0.5510, device='cuda:0')...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3595</th>\n",
              "      <td>NaN</td>\n",
              "      <td>13.0</td>\n",
              "      <td>model:net_state_dict</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>{'0.weight': [[tensor(0.5510, device='cuda:0')...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3852</th>\n",
              "      <td>NaN</td>\n",
              "      <td>14.0</td>\n",
              "      <td>model:net_state_dict</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>{'0.weight': [[tensor(0.5510, device='cuda:0')...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4109</th>\n",
              "      <td>NaN</td>\n",
              "      <td>15.0</td>\n",
              "      <td>model:net_state_dict</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>{'0.weight': [[tensor(0.5510, device='cuda:0')...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4366</th>\n",
              "      <td>NaN</td>\n",
              "      <td>16.0</td>\n",
              "      <td>model:net_state_dict</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>{'0.weight': [[tensor(0.5510, device='cuda:0')...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4623</th>\n",
              "      <td>NaN</td>\n",
              "      <td>17.0</td>\n",
              "      <td>model:net_state_dict</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>{'0.weight': [[tensor(0.5510, device='cuda:0')...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4880</th>\n",
              "      <td>NaN</td>\n",
              "      <td>18.0</td>\n",
              "      <td>model:net_state_dict</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>{'0.weight': [[tensor(0.5510, device='cuda:0')...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5137</th>\n",
              "      <td>NaN</td>\n",
              "      <td>19.0</td>\n",
              "      <td>model:net_state_dict</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>{'0.weight': [[tensor(0.5510, device='cuda:0')...</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "      i_batch  ...                                              value\n",
              "255       NaN  ...  {'0.weight': [[tensor(0.5510, device='cuda:0')...\n",
              "512       NaN  ...  {'0.weight': [[tensor(0.5510, device='cuda:0')...\n",
              "769       NaN  ...  {'0.weight': [[tensor(0.5510, device='cuda:0')...\n",
              "1026      NaN  ...  {'0.weight': [[tensor(0.5510, device='cuda:0')...\n",
              "1283      NaN  ...  {'0.weight': [[tensor(0.5510, device='cuda:0')...\n",
              "1540      NaN  ...  {'0.weight': [[tensor(0.5510, device='cuda:0')...\n",
              "1797      NaN  ...  {'0.weight': [[tensor(0.5510, device='cuda:0')...\n",
              "2054      NaN  ...  {'0.weight': [[tensor(0.5510, device='cuda:0')...\n",
              "2311      NaN  ...  {'0.weight': [[tensor(0.5510, device='cuda:0')...\n",
              "2567      NaN  ...  {'0.weight': [[tensor(0.5510, device='cuda:0')...\n",
              "2824      NaN  ...  {'0.weight': [[tensor(0.5510, device='cuda:0')...\n",
              "3081      NaN  ...  {'0.weight': [[tensor(0.5510, device='cuda:0')...\n",
              "3338      NaN  ...  {'0.weight': [[tensor(0.5510, device='cuda:0')...\n",
              "3595      NaN  ...  {'0.weight': [[tensor(0.5510, device='cuda:0')...\n",
              "3852      NaN  ...  {'0.weight': [[tensor(0.5510, device='cuda:0')...\n",
              "4109      NaN  ...  {'0.weight': [[tensor(0.5510, device='cuda:0')...\n",
              "4366      NaN  ...  {'0.weight': [[tensor(0.5510, device='cuda:0')...\n",
              "4623      NaN  ...  {'0.weight': [[tensor(0.5510, device='cuda:0')...\n",
              "4880      NaN  ...  {'0.weight': [[tensor(0.5510, device='cuda:0')...\n",
              "5137      NaN  ...  {'0.weight': [[tensor(0.5510, device='cuda:0')...\n",
              "\n",
              "[20 rows x 5 columns]"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "per_epoch optim:adam_state_dict\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>i_batch</th>\n",
              "      <th>i_epoch</th>\n",
              "      <th>key</th>\n",
              "      <th>label</th>\n",
              "      <th>value</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>256</th>\n",
              "      <td>NaN</td>\n",
              "      <td>0.0</td>\n",
              "      <td>optim:adam_state_dict</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>{'state': {0: {'step': 600, 'exp_avg': tensor(...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>513</th>\n",
              "      <td>NaN</td>\n",
              "      <td>1.0</td>\n",
              "      <td>optim:adam_state_dict</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>{'state': {0: {'step': 600, 'exp_avg': tensor(...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>770</th>\n",
              "      <td>NaN</td>\n",
              "      <td>2.0</td>\n",
              "      <td>optim:adam_state_dict</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>{'state': {0: {'step': 600, 'exp_avg': tensor(...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1027</th>\n",
              "      <td>NaN</td>\n",
              "      <td>3.0</td>\n",
              "      <td>optim:adam_state_dict</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>{'state': {0: {'step': 600, 'exp_avg': tensor(...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1284</th>\n",
              "      <td>NaN</td>\n",
              "      <td>4.0</td>\n",
              "      <td>optim:adam_state_dict</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>{'state': {0: {'step': 600, 'exp_avg': tensor(...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1541</th>\n",
              "      <td>NaN</td>\n",
              "      <td>5.0</td>\n",
              "      <td>optim:adam_state_dict</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>{'state': {0: {'step': 600, 'exp_avg': tensor(...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1798</th>\n",
              "      <td>NaN</td>\n",
              "      <td>6.0</td>\n",
              "      <td>optim:adam_state_dict</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>{'state': {0: {'step': 600, 'exp_avg': tensor(...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2055</th>\n",
              "      <td>NaN</td>\n",
              "      <td>7.0</td>\n",
              "      <td>optim:adam_state_dict</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>{'state': {0: {'step': 600, 'exp_avg': tensor(...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2312</th>\n",
              "      <td>NaN</td>\n",
              "      <td>8.0</td>\n",
              "      <td>optim:adam_state_dict</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>{'state': {0: {'step': 600, 'exp_avg': tensor(...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2568</th>\n",
              "      <td>NaN</td>\n",
              "      <td>9.0</td>\n",
              "      <td>optim:adam_state_dict</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>{'state': {0: {'step': 600, 'exp_avg': tensor(...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2825</th>\n",
              "      <td>NaN</td>\n",
              "      <td>10.0</td>\n",
              "      <td>optim:adam_state_dict</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>{'state': {0: {'step': 600, 'exp_avg': tensor(...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3082</th>\n",
              "      <td>NaN</td>\n",
              "      <td>11.0</td>\n",
              "      <td>optim:adam_state_dict</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>{'state': {0: {'step': 600, 'exp_avg': tensor(...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3339</th>\n",
              "      <td>NaN</td>\n",
              "      <td>12.0</td>\n",
              "      <td>optim:adam_state_dict</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>{'state': {0: {'step': 600, 'exp_avg': tensor(...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3596</th>\n",
              "      <td>NaN</td>\n",
              "      <td>13.0</td>\n",
              "      <td>optim:adam_state_dict</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>{'state': {0: {'step': 600, 'exp_avg': tensor(...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3853</th>\n",
              "      <td>NaN</td>\n",
              "      <td>14.0</td>\n",
              "      <td>optim:adam_state_dict</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>{'state': {0: {'step': 600, 'exp_avg': tensor(...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4110</th>\n",
              "      <td>NaN</td>\n",
              "      <td>15.0</td>\n",
              "      <td>optim:adam_state_dict</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>{'state': {0: {'step': 600, 'exp_avg': tensor(...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4367</th>\n",
              "      <td>NaN</td>\n",
              "      <td>16.0</td>\n",
              "      <td>optim:adam_state_dict</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>{'state': {0: {'step': 600, 'exp_avg': tensor(...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4624</th>\n",
              "      <td>NaN</td>\n",
              "      <td>17.0</td>\n",
              "      <td>optim:adam_state_dict</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>{'state': {0: {'step': 600, 'exp_avg': tensor(...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4881</th>\n",
              "      <td>NaN</td>\n",
              "      <td>18.0</td>\n",
              "      <td>optim:adam_state_dict</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>{'state': {0: {'step': 600, 'exp_avg': tensor(...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5138</th>\n",
              "      <td>NaN</td>\n",
              "      <td>19.0</td>\n",
              "      <td>optim:adam_state_dict</td>\n",
              "      <td>per_epoch</td>\n",
              "      <td>{'state': {0: {'step': 600, 'exp_avg': tensor(...</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "      i_batch  ...                                              value\n",
              "256       NaN  ...  {'state': {0: {'step': 600, 'exp_avg': tensor(...\n",
              "513       NaN  ...  {'state': {0: {'step': 600, 'exp_avg': tensor(...\n",
              "770       NaN  ...  {'state': {0: {'step': 600, 'exp_avg': tensor(...\n",
              "1027      NaN  ...  {'state': {0: {'step': 600, 'exp_avg': tensor(...\n",
              "1284      NaN  ...  {'state': {0: {'step': 600, 'exp_avg': tensor(...\n",
              "1541      NaN  ...  {'state': {0: {'step': 600, 'exp_avg': tensor(...\n",
              "1798      NaN  ...  {'state': {0: {'step': 600, 'exp_avg': tensor(...\n",
              "2055      NaN  ...  {'state': {0: {'step': 600, 'exp_avg': tensor(...\n",
              "2312      NaN  ...  {'state': {0: {'step': 600, 'exp_avg': tensor(...\n",
              "2568      NaN  ...  {'state': {0: {'step': 600, 'exp_avg': tensor(...\n",
              "2825      NaN  ...  {'state': {0: {'step': 600, 'exp_avg': tensor(...\n",
              "3082      NaN  ...  {'state': {0: {'step': 600, 'exp_avg': tensor(...\n",
              "3339      NaN  ...  {'state': {0: {'step': 600, 'exp_avg': tensor(...\n",
              "3596      NaN  ...  {'state': {0: {'step': 600, 'exp_avg': tensor(...\n",
              "3853      NaN  ...  {'state': {0: {'step': 600, 'exp_avg': tensor(...\n",
              "4110      NaN  ...  {'state': {0: {'step': 600, 'exp_avg': tensor(...\n",
              "4367      NaN  ...  {'state': {0: {'step': 600, 'exp_avg': tensor(...\n",
              "4624      NaN  ...  {'state': {0: {'step': 600, 'exp_avg': tensor(...\n",
              "4881      NaN  ...  {'state': {0: {'step': 600, 'exp_avg': tensor(...\n",
              "5138      NaN  ...  {'state': {0: {'step': 600, 'exp_avg': tensor(...\n",
              "\n",
              "[20 rows x 5 columns]"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "per_batch time:elapsed\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>i_batch</th>\n",
              "      <th>i_epoch</th>\n",
              "      <th>key</th>\n",
              "      <th>label</th>\n",
              "      <th>value</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>6</th>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>time:elapsed</td>\n",
              "      <td>per_batch</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>14</th>\n",
              "      <td>1.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>time:elapsed</td>\n",
              "      <td>per_batch</td>\n",
              "      <td>0.0342147</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>22</th>\n",
              "      <td>2.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>time:elapsed</td>\n",
              "      <td>per_batch</td>\n",
              "      <td>0.0653546</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>30</th>\n",
              "      <td>3.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>time:elapsed</td>\n",
              "      <td>per_batch</td>\n",
              "      <td>0.0966768</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>38</th>\n",
              "      <td>4.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>time:elapsed</td>\n",
              "      <td>per_batch</td>\n",
              "      <td>0.127857</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5088</th>\n",
              "      <td>25.0</td>\n",
              "      <td>19.0</td>\n",
              "      <td>time:elapsed</td>\n",
              "      <td>per_batch</td>\n",
              "      <td>0.863674</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5096</th>\n",
              "      <td>26.0</td>\n",
              "      <td>19.0</td>\n",
              "      <td>time:elapsed</td>\n",
              "      <td>per_batch</td>\n",
              "      <td>0.897095</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5104</th>\n",
              "      <td>27.0</td>\n",
              "      <td>19.0</td>\n",
              "      <td>time:elapsed</td>\n",
              "      <td>per_batch</td>\n",
              "      <td>0.930181</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5112</th>\n",
              "      <td>28.0</td>\n",
              "      <td>19.0</td>\n",
              "      <td>time:elapsed</td>\n",
              "      <td>per_batch</td>\n",
              "      <td>0.963495</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5120</th>\n",
              "      <td>29.0</td>\n",
              "      <td>19.0</td>\n",
              "      <td>time:elapsed</td>\n",
              "      <td>per_batch</td>\n",
              "      <td>0.996997</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>600 rows × 5 columns</p>\n",
              "</div>"
            ],
            "text/plain": [
              "      i_batch  i_epoch           key      label      value\n",
              "6         0.0      0.0  time:elapsed  per_batch          0\n",
              "14        1.0      0.0  time:elapsed  per_batch  0.0342147\n",
              "22        2.0      0.0  time:elapsed  per_batch  0.0653546\n",
              "30        3.0      0.0  time:elapsed  per_batch  0.0966768\n",
              "38        4.0      0.0  time:elapsed  per_batch   0.127857\n",
              "...       ...      ...           ...        ...        ...\n",
              "5088     25.0     19.0  time:elapsed  per_batch   0.863674\n",
              "5096     26.0     19.0  time:elapsed  per_batch   0.897095\n",
              "5104     27.0     19.0  time:elapsed  per_batch   0.930181\n",
              "5112     28.0     19.0  time:elapsed  per_batch   0.963495\n",
              "5120     29.0     19.0  time:elapsed  per_batch   0.996997\n",
              "\n",
              "[600 rows x 5 columns]"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "per_batch time:total\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>i_batch</th>\n",
              "      <th>i_epoch</th>\n",
              "      <th>key</th>\n",
              "      <th>label</th>\n",
              "      <th>value</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>7</th>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>time:total</td>\n",
              "      <td>per_batch</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>15</th>\n",
              "      <td>1.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>time:total</td>\n",
              "      <td>per_batch</td>\n",
              "      <td>1.02644</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>23</th>\n",
              "      <td>2.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>time:total</td>\n",
              "      <td>per_batch</td>\n",
              "      <td>0.980319</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>31</th>\n",
              "      <td>3.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>time:total</td>\n",
              "      <td>per_batch</td>\n",
              "      <td>0.966768</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>39</th>\n",
              "      <td>4.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>time:total</td>\n",
              "      <td>per_batch</td>\n",
              "      <td>0.958931</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5089</th>\n",
              "      <td>25.0</td>\n",
              "      <td>19.0</td>\n",
              "      <td>time:total</td>\n",
              "      <td>per_batch</td>\n",
              "      <td>1.03641</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5097</th>\n",
              "      <td>26.0</td>\n",
              "      <td>19.0</td>\n",
              "      <td>time:total</td>\n",
              "      <td>per_batch</td>\n",
              "      <td>1.03511</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5105</th>\n",
              "      <td>27.0</td>\n",
              "      <td>19.0</td>\n",
              "      <td>time:total</td>\n",
              "      <td>per_batch</td>\n",
              "      <td>1.03353</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5113</th>\n",
              "      <td>28.0</td>\n",
              "      <td>19.0</td>\n",
              "      <td>time:total</td>\n",
              "      <td>per_batch</td>\n",
              "      <td>1.03232</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5121</th>\n",
              "      <td>29.0</td>\n",
              "      <td>19.0</td>\n",
              "      <td>time:total</td>\n",
              "      <td>per_batch</td>\n",
              "      <td>1.03138</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>600 rows × 5 columns</p>\n",
              "</div>"
            ],
            "text/plain": [
              "      i_batch  i_epoch         key      label     value\n",
              "7         0.0      0.0  time:total  per_batch         0\n",
              "15        1.0      0.0  time:total  per_batch   1.02644\n",
              "23        2.0      0.0  time:total  per_batch  0.980319\n",
              "31        3.0      0.0  time:total  per_batch  0.966768\n",
              "39        4.0      0.0  time:total  per_batch  0.958931\n",
              "...       ...      ...         ...        ...       ...\n",
              "5089     25.0     19.0  time:total  per_batch   1.03641\n",
              "5097     26.0     19.0  time:total  per_batch   1.03511\n",
              "5105     27.0     19.0  time:total  per_batch   1.03353\n",
              "5113     28.0     19.0  time:total  per_batch   1.03232\n",
              "5121     29.0     19.0  time:total  per_batch   1.03138\n",
              "\n",
              "[600 rows x 5 columns]"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "per_batch time:left\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>i_batch</th>\n",
              "      <th>i_epoch</th>\n",
              "      <th>key</th>\n",
              "      <th>label</th>\n",
              "      <th>value</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>8</th>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>time:left</td>\n",
              "      <td>per_batch</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>16</th>\n",
              "      <td>1.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>time:left</td>\n",
              "      <td>per_batch</td>\n",
              "      <td>0.992227</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>24</th>\n",
              "      <td>2.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>time:left</td>\n",
              "      <td>per_batch</td>\n",
              "      <td>0.914964</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>32</th>\n",
              "      <td>3.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>time:left</td>\n",
              "      <td>per_batch</td>\n",
              "      <td>0.870091</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>40</th>\n",
              "      <td>4.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>time:left</td>\n",
              "      <td>per_batch</td>\n",
              "      <td>0.831073</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5090</th>\n",
              "      <td>25.0</td>\n",
              "      <td>19.0</td>\n",
              "      <td>time:left</td>\n",
              "      <td>per_batch</td>\n",
              "      <td>0.172735</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5098</th>\n",
              "      <td>26.0</td>\n",
              "      <td>19.0</td>\n",
              "      <td>time:left</td>\n",
              "      <td>per_batch</td>\n",
              "      <td>0.138015</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5106</th>\n",
              "      <td>27.0</td>\n",
              "      <td>19.0</td>\n",
              "      <td>time:left</td>\n",
              "      <td>per_batch</td>\n",
              "      <td>0.103353</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5114</th>\n",
              "      <td>28.0</td>\n",
              "      <td>19.0</td>\n",
              "      <td>time:left</td>\n",
              "      <td>per_batch</td>\n",
              "      <td>0.0688211</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5122</th>\n",
              "      <td>29.0</td>\n",
              "      <td>19.0</td>\n",
              "      <td>time:left</td>\n",
              "      <td>per_batch</td>\n",
              "      <td>0.0343792</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>600 rows × 5 columns</p>\n",
              "</div>"
            ],
            "text/plain": [
              "      i_batch  i_epoch        key      label      value\n",
              "8         0.0      0.0  time:left  per_batch          0\n",
              "16        1.0      0.0  time:left  per_batch   0.992227\n",
              "24        2.0      0.0  time:left  per_batch   0.914964\n",
              "32        3.0      0.0  time:left  per_batch   0.870091\n",
              "40        4.0      0.0  time:left  per_batch   0.831073\n",
              "...       ...      ...        ...        ...        ...\n",
              "5090     25.0     19.0  time:left  per_batch   0.172735\n",
              "5098     26.0     19.0  time:left  per_batch   0.138015\n",
              "5106     27.0     19.0  time:left  per_batch   0.103353\n",
              "5114     28.0     19.0  time:left  per_batch  0.0688211\n",
              "5122     29.0     19.0  time:left  per_batch  0.0343792\n",
              "\n",
              "[600 rows x 5 columns]"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "per_batch time:each\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>i_batch</th>\n",
              "      <th>i_epoch</th>\n",
              "      <th>key</th>\n",
              "      <th>label</th>\n",
              "      <th>value</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>9</th>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>time:each</td>\n",
              "      <td>per_batch</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>17</th>\n",
              "      <td>1.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>time:each</td>\n",
              "      <td>per_batch</td>\n",
              "      <td>0.0342147</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>25</th>\n",
              "      <td>2.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>time:each</td>\n",
              "      <td>per_batch</td>\n",
              "      <td>0.0326773</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>33</th>\n",
              "      <td>3.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>time:each</td>\n",
              "      <td>per_batch</td>\n",
              "      <td>0.0322256</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>41</th>\n",
              "      <td>4.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>time:each</td>\n",
              "      <td>per_batch</td>\n",
              "      <td>0.0319644</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5091</th>\n",
              "      <td>25.0</td>\n",
              "      <td>19.0</td>\n",
              "      <td>time:each</td>\n",
              "      <td>per_batch</td>\n",
              "      <td>0.034547</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5099</th>\n",
              "      <td>26.0</td>\n",
              "      <td>19.0</td>\n",
              "      <td>time:each</td>\n",
              "      <td>per_batch</td>\n",
              "      <td>0.0345036</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5107</th>\n",
              "      <td>27.0</td>\n",
              "      <td>19.0</td>\n",
              "      <td>time:each</td>\n",
              "      <td>per_batch</td>\n",
              "      <td>0.0344511</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5115</th>\n",
              "      <td>28.0</td>\n",
              "      <td>19.0</td>\n",
              "      <td>time:each</td>\n",
              "      <td>per_batch</td>\n",
              "      <td>0.0344105</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5123</th>\n",
              "      <td>29.0</td>\n",
              "      <td>19.0</td>\n",
              "      <td>time:each</td>\n",
              "      <td>per_batch</td>\n",
              "      <td>0.0343792</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>600 rows × 5 columns</p>\n",
              "</div>"
            ],
            "text/plain": [
              "      i_batch  i_epoch        key      label      value\n",
              "9         0.0      0.0  time:each  per_batch          0\n",
              "17        1.0      0.0  time:each  per_batch  0.0342147\n",
              "25        2.0      0.0  time:each  per_batch  0.0326773\n",
              "33        3.0      0.0  time:each  per_batch  0.0322256\n",
              "41        4.0      0.0  time:each  per_batch  0.0319644\n",
              "...       ...      ...        ...        ...        ...\n",
              "5091     25.0     19.0  time:each  per_batch   0.034547\n",
              "5099     26.0     19.0  time:each  per_batch  0.0345036\n",
              "5107     27.0     19.0  time:each  per_batch  0.0344511\n",
              "5115     28.0     19.0  time:each  per_batch  0.0344105\n",
              "5123     29.0     19.0  time:each  per_batch  0.0343792\n",
              "\n",
              "[600 rows x 5 columns]"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "per_batch loss:mae\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>i_batch</th>\n",
              "      <th>i_epoch</th>\n",
              "      <th>key</th>\n",
              "      <th>label</th>\n",
              "      <th>value</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>10</th>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>loss:mae</td>\n",
              "      <td>per_batch</td>\n",
              "      <td>1.39187</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>18</th>\n",
              "      <td>1.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>loss:mae</td>\n",
              "      <td>per_batch</td>\n",
              "      <td>0.799705</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>26</th>\n",
              "      <td>2.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>loss:mae</td>\n",
              "      <td>per_batch</td>\n",
              "      <td>1.15159</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>34</th>\n",
              "      <td>3.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>loss:mae</td>\n",
              "      <td>per_batch</td>\n",
              "      <td>1.21475</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>42</th>\n",
              "      <td>4.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>loss:mae</td>\n",
              "      <td>per_batch</td>\n",
              "      <td>0.851634</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5092</th>\n",
              "      <td>25.0</td>\n",
              "      <td>19.0</td>\n",
              "      <td>loss:mae</td>\n",
              "      <td>per_batch</td>\n",
              "      <td>0.00199799</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5100</th>\n",
              "      <td>26.0</td>\n",
              "      <td>19.0</td>\n",
              "      <td>loss:mae</td>\n",
              "      <td>per_batch</td>\n",
              "      <td>0.0020025</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5108</th>\n",
              "      <td>27.0</td>\n",
              "      <td>19.0</td>\n",
              "      <td>loss:mae</td>\n",
              "      <td>per_batch</td>\n",
              "      <td>0.00201816</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5116</th>\n",
              "      <td>28.0</td>\n",
              "      <td>19.0</td>\n",
              "      <td>loss:mae</td>\n",
              "      <td>per_batch</td>\n",
              "      <td>0.00202666</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5124</th>\n",
              "      <td>29.0</td>\n",
              "      <td>19.0</td>\n",
              "      <td>loss:mae</td>\n",
              "      <td>per_batch</td>\n",
              "      <td>0.00200885</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>600 rows × 5 columns</p>\n",
              "</div>"
            ],
            "text/plain": [
              "      i_batch  i_epoch       key      label       value\n",
              "10        0.0      0.0  loss:mae  per_batch     1.39187\n",
              "18        1.0      0.0  loss:mae  per_batch    0.799705\n",
              "26        2.0      0.0  loss:mae  per_batch     1.15159\n",
              "34        3.0      0.0  loss:mae  per_batch     1.21475\n",
              "42        4.0      0.0  loss:mae  per_batch    0.851634\n",
              "...       ...      ...       ...        ...         ...\n",
              "5092     25.0     19.0  loss:mae  per_batch  0.00199799\n",
              "5100     26.0     19.0  loss:mae  per_batch   0.0020025\n",
              "5108     27.0     19.0  loss:mae  per_batch  0.00201816\n",
              "5116     28.0     19.0  loss:mae  per_batch  0.00202666\n",
              "5124     29.0     19.0  loss:mae  per_batch  0.00200885\n",
              "\n",
              "[600 rows x 5 columns]"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "per_batch loss:mse\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>i_batch</th>\n",
              "      <th>i_epoch</th>\n",
              "      <th>key</th>\n",
              "      <th>label</th>\n",
              "      <th>value</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>11</th>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>loss:mse</td>\n",
              "      <td>per_batch</td>\n",
              "      <td>3.57444</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>19</th>\n",
              "      <td>1.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>loss:mse</td>\n",
              "      <td>per_batch</td>\n",
              "      <td>0.84209</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>27</th>\n",
              "      <td>2.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>loss:mse</td>\n",
              "      <td>per_batch</td>\n",
              "      <td>1.66108</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>35</th>\n",
              "      <td>3.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>loss:mse</td>\n",
              "      <td>per_batch</td>\n",
              "      <td>1.73666</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>43</th>\n",
              "      <td>4.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>loss:mse</td>\n",
              "      <td>per_batch</td>\n",
              "      <td>0.930269</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5093</th>\n",
              "      <td>25.0</td>\n",
              "      <td>19.0</td>\n",
              "      <td>loss:mse</td>\n",
              "      <td>per_batch</td>\n",
              "      <td>7.47617e-06</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5101</th>\n",
              "      <td>26.0</td>\n",
              "      <td>19.0</td>\n",
              "      <td>loss:mse</td>\n",
              "      <td>per_batch</td>\n",
              "      <td>7.34517e-06</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5109</th>\n",
              "      <td>27.0</td>\n",
              "      <td>19.0</td>\n",
              "      <td>loss:mse</td>\n",
              "      <td>per_batch</td>\n",
              "      <td>7.97422e-06</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5117</th>\n",
              "      <td>28.0</td>\n",
              "      <td>19.0</td>\n",
              "      <td>loss:mse</td>\n",
              "      <td>per_batch</td>\n",
              "      <td>7.95203e-06</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5125</th>\n",
              "      <td>29.0</td>\n",
              "      <td>19.0</td>\n",
              "      <td>loss:mse</td>\n",
              "      <td>per_batch</td>\n",
              "      <td>7.71786e-06</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>600 rows × 5 columns</p>\n",
              "</div>"
            ],
            "text/plain": [
              "      i_batch  i_epoch       key      label        value\n",
              "11        0.0      0.0  loss:mse  per_batch      3.57444\n",
              "19        1.0      0.0  loss:mse  per_batch      0.84209\n",
              "27        2.0      0.0  loss:mse  per_batch      1.66108\n",
              "35        3.0      0.0  loss:mse  per_batch      1.73666\n",
              "43        4.0      0.0  loss:mse  per_batch     0.930269\n",
              "...       ...      ...       ...        ...          ...\n",
              "5093     25.0     19.0  loss:mse  per_batch  7.47617e-06\n",
              "5101     26.0     19.0  loss:mse  per_batch  7.34517e-06\n",
              "5109     27.0     19.0  loss:mse  per_batch  7.97422e-06\n",
              "5117     28.0     19.0  loss:mse  per_batch  7.95203e-06\n",
              "5125     29.0     19.0  loss:mse  per_batch  7.71786e-06\n",
              "\n",
              "[600 rows x 5 columns]"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "per_batch metric:1\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>i_batch</th>\n",
              "      <th>i_epoch</th>\n",
              "      <th>key</th>\n",
              "      <th>label</th>\n",
              "      <th>value</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>12</th>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>metric:1</td>\n",
              "      <td>per_batch</td>\n",
              "      <td>4.96631</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>20</th>\n",
              "      <td>1.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>metric:1</td>\n",
              "      <td>per_batch</td>\n",
              "      <td>1.64179</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>28</th>\n",
              "      <td>2.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>metric:1</td>\n",
              "      <td>per_batch</td>\n",
              "      <td>2.81267</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>36</th>\n",
              "      <td>3.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>metric:1</td>\n",
              "      <td>per_batch</td>\n",
              "      <td>2.95141</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>44</th>\n",
              "      <td>4.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>metric:1</td>\n",
              "      <td>per_batch</td>\n",
              "      <td>1.7819</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5094</th>\n",
              "      <td>25.0</td>\n",
              "      <td>19.0</td>\n",
              "      <td>metric:1</td>\n",
              "      <td>per_batch</td>\n",
              "      <td>0.00200547</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5102</th>\n",
              "      <td>26.0</td>\n",
              "      <td>19.0</td>\n",
              "      <td>metric:1</td>\n",
              "      <td>per_batch</td>\n",
              "      <td>0.00200984</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5110</th>\n",
              "      <td>27.0</td>\n",
              "      <td>19.0</td>\n",
              "      <td>metric:1</td>\n",
              "      <td>per_batch</td>\n",
              "      <td>0.00202613</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5118</th>\n",
              "      <td>28.0</td>\n",
              "      <td>19.0</td>\n",
              "      <td>metric:1</td>\n",
              "      <td>per_batch</td>\n",
              "      <td>0.00203461</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5126</th>\n",
              "      <td>29.0</td>\n",
              "      <td>19.0</td>\n",
              "      <td>metric:1</td>\n",
              "      <td>per_batch</td>\n",
              "      <td>0.00201657</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>600 rows × 5 columns</p>\n",
              "</div>"
            ],
            "text/plain": [
              "      i_batch  i_epoch       key      label       value\n",
              "12        0.0      0.0  metric:1  per_batch     4.96631\n",
              "20        1.0      0.0  metric:1  per_batch     1.64179\n",
              "28        2.0      0.0  metric:1  per_batch     2.81267\n",
              "36        3.0      0.0  metric:1  per_batch     2.95141\n",
              "44        4.0      0.0  metric:1  per_batch      1.7819\n",
              "...       ...      ...       ...        ...         ...\n",
              "5094     25.0     19.0  metric:1  per_batch  0.00200547\n",
              "5102     26.0     19.0  metric:1  per_batch  0.00200984\n",
              "5110     27.0     19.0  metric:1  per_batch  0.00202613\n",
              "5118     28.0     19.0  metric:1  per_batch  0.00203461\n",
              "5126     29.0     19.0  metric:1  per_batch  0.00201657\n",
              "\n",
              "[600 rows x 5 columns]"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "per_batch metric:2\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>i_batch</th>\n",
              "      <th>i_epoch</th>\n",
              "      <th>key</th>\n",
              "      <th>label</th>\n",
              "      <th>value</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>13</th>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>metric:2</td>\n",
              "      <td>per_batch</td>\n",
              "      <td>-2.18257</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>21</th>\n",
              "      <td>1.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>metric:2</td>\n",
              "      <td>per_batch</td>\n",
              "      <td>-0.0423852</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>29</th>\n",
              "      <td>2.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>metric:2</td>\n",
              "      <td>per_batch</td>\n",
              "      <td>-0.509488</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>37</th>\n",
              "      <td>3.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>metric:2</td>\n",
              "      <td>per_batch</td>\n",
              "      <td>-0.521905</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>45</th>\n",
              "      <td>4.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>metric:2</td>\n",
              "      <td>per_batch</td>\n",
              "      <td>-0.0786344</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5095</th>\n",
              "      <td>25.0</td>\n",
              "      <td>19.0</td>\n",
              "      <td>metric:2</td>\n",
              "      <td>per_batch</td>\n",
              "      <td>0.00199052</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5103</th>\n",
              "      <td>26.0</td>\n",
              "      <td>19.0</td>\n",
              "      <td>metric:2</td>\n",
              "      <td>per_batch</td>\n",
              "      <td>0.00199515</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5111</th>\n",
              "      <td>27.0</td>\n",
              "      <td>19.0</td>\n",
              "      <td>metric:2</td>\n",
              "      <td>per_batch</td>\n",
              "      <td>0.00201018</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5119</th>\n",
              "      <td>28.0</td>\n",
              "      <td>19.0</td>\n",
              "      <td>metric:2</td>\n",
              "      <td>per_batch</td>\n",
              "      <td>0.00201871</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5127</th>\n",
              "      <td>29.0</td>\n",
              "      <td>19.0</td>\n",
              "      <td>metric:2</td>\n",
              "      <td>per_batch</td>\n",
              "      <td>0.00200113</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>600 rows × 5 columns</p>\n",
              "</div>"
            ],
            "text/plain": [
              "      i_batch  i_epoch       key      label       value\n",
              "13        0.0      0.0  metric:2  per_batch    -2.18257\n",
              "21        1.0      0.0  metric:2  per_batch  -0.0423852\n",
              "29        2.0      0.0  metric:2  per_batch   -0.509488\n",
              "37        3.0      0.0  metric:2  per_batch   -0.521905\n",
              "45        4.0      0.0  metric:2  per_batch  -0.0786344\n",
              "...       ...      ...       ...        ...         ...\n",
              "5095     25.0     19.0  metric:2  per_batch  0.00199052\n",
              "5103     26.0     19.0  metric:2  per_batch  0.00199515\n",
              "5111     27.0     19.0  metric:2  per_batch  0.00201018\n",
              "5119     28.0     19.0  metric:2  per_batch  0.00201871\n",
              "5127     29.0     19.0  metric:2  per_batch  0.00200113\n",
              "\n",
              "[600 rows x 5 columns]"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "bASaZ1B1s-Pt",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 623
        },
        "outputId": "445f6eb3-6585-4d78-cb71-b5e9b96319cc"
      },
      "source": [
        "i_label = 'per_epoch'\n",
        "i_key_group = 'loss'\n",
        "\n",
        "plt.figure(figsize=[20,10])\n",
        "plt.xlabel('i_epoch')\n",
        "plt.ylabel('MinMax scaled values')\n",
        "for ik in enb2.dflog['key'].loc[enb2.dflog['label']==i_label].unique():\n",
        "    if ik[:len(i_key_group)] != i_key_group: continue\n",
        "    tdf = enb2.dflog.loc[(enb2.dflog['key']==ik) & (enb2.dflog['label']==i_label)]\n",
        "    plt.semilogy(tdf['i_epoch'],(tdf['value']-tdf['value'].min())/(tdf['value'].max()-tdf['value'].min()),label=ik)\n",
        "plt.legend(framealpha=0.7)"
      ],
      "execution_count": 12,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.legend.Legend at 0x7f89260ae9b0>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 12
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1440x720 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "RNJgsAhyVDd1",
        "colab_type": "text"
      },
      "source": [
        "# scratch"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "fgO5qQnbVFM6",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "spam = torch.nn.Parameter(torch.randn(1).to(d))"
      ],
      "execution_count": 13,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "HK3fCF-rVIrp",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 136
        },
        "outputId": "ce1dfad3-c2da-4a2c-fb18-8eef2dcb4d3f"
      },
      "source": [
        "print('',spam)\n",
        "print('.detach()',spam.detach())\n",
        "print('.cpu()',spam.cpu())\n",
        "print('.item()',spam.item())\n",
        "print('.detach().cpu()',spam.detach().cpu())\n",
        "print('.detach().cpu().item()',spam.detach().cpu().item())"
      ],
      "execution_count": 14,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            " Parameter containing:\n",
            "tensor([-1.9656], device='cuda:0', requires_grad=True)\n",
            ".detach() tensor([-1.9656], device='cuda:0')\n",
            ".cpu() tensor([-1.9656], grad_fn=<CopyBackwards>)\n",
            ".item() -1.9656450748443604\n",
            ".detach().cpu() tensor([-1.9656])\n",
            ".detach().cpu().item() -1.9656450748443604\n"
          ],
          "name": "stdout"
        }
      ]
    }
  ]
}